Steve Telford wrote:
>
> Regarding PIM Sparse, Dense and Sparse-Dense modes, does anyone
> know why the
> often stated design preference for sparse-dense exists?
I think that the logic is that with sparese-dense, the m-cast network could
continue to function even if the RP were to die. It's just
Teflon is correct. If you configure your network to use Auto-RP, then all
of your interfaces will have to use PIM Sparse-Dense Mode. Besides the fact
that Auto-RP uses dense mode for it's control data, if any router looses
connectivity to the RP, then it would fall back to flood-and-prune. To me
Thanks for the info Doan
Is this info from experience of large multicast deployment?
What also I was trying to get at is the question of whether with the
enhancement of auto-rp listener is the need for sparse-dense totally negated?
This would be regardless of the overhead issues of using auto-rp
Steve,
I'm working on multicast design a new DoD network. At this level we are a
pretty big network with OC192 pipes. At this level I rather not use any
dense mode type of implementation with my multicast design if possible. Got
too many other protocols to deal with already. I'm currently at C
4 matches
Mail list logo
