From: Jian Gu (guxiaojian@gmail.com)
Date: Tue Feb 06 2007 - 15:36:32 ART
On 2/6/07, johngibson1541@yahoo.com <johngibson1541@yahoo.com> wrote:
>
> I know (S,G) will still exist in the path from the source to RP even in
> strict sparse-mode.
What do you mean "(S,G) still exist in the path from the source to RP even
in strict sparse-mode"? in sparse mode, RP explicitly sends (S,G) join to
the source. (*,G) entry should not be used for multicast forwarding in SPT.
In my home lab, only (*,G) is populated from the destination to the RP
> before I actually send ping, everything seems correct until I actually sent
> ping from source to destination.
>
> Once I actually send ping from the source to the destination. (S,G) just
> shows up in every router along the path.
>
> After the ping. If I wait for a few minutes, the extra (S,G) will
> disappear, everything comes back to normal.
>
> Is this normal ?
My doubt is that if (S,G) will be everywhere when the multicast application
> is sending packets, then sparse-mode is not superior than dense-mode at all
> if only consider keeping the mroute table small.
PIM-bidir will have less mroutes, the reason sparse-mode is superior than
dense is because the protocol does not flood multicast traffic.
So, in concern of keeping the mroute table small, sparse-mode is superior
> only if the applications don't constantly send data.
>
>
> John
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Thu Mar 01 2007 - 07:38:46 ART