From: Scott Smith (hioctane@gmail.com)
Date: Wed Jan 25 2006 - 17:18:31 GMT-3
Hi,
I've got a multicast issue with this lab and I'm not sure how to fix
it. PIM sparse mode with static RP is the setup.
R2 is the RP. The problem is with the join coming from a host behind
R1. R1 forwards the join to R3 per the routing table out s0/0 to R3.
The join arrives at R3 s1/0. Now here is where things get interesting.
The routing table would have R3 send the join to R2 via s1/3 but pim
isn't running on that interface so what appears to happen is R3 sends
the join to the bit bucket and therefore the join never occurs. This
isn't an RPF failure (if it is I don't see how). We're going upstream
to the RP. All the RPF checks to this point are ok. The solution
mentions a fix to the RPF failure that would occur on R3 if the join
could complete.
So, the way I understand this... for a router to find the path to the
RP it consults the routing table then sends the join request to the RP
along that path/int. So what happens when the preferred exit point to
the RP isn't running PIM?
All this fixes I've come up with violate the rules of the lab. Of
course I could be totally confused and this may not actually be a
problem:-)
TIA!
-Scott
This archive was generated by hypermail 2.1.4 : Wed Feb 01 2006 - 07:45:50 GMT-3