From: Frank B (frank@xxxxxxxxxxxx)
Date: Sun Aug 04 2002 - 00:43:47 GMT-3
Hey Tony and Jeff,
I don't notice anything unusual except that you're using
interface e1 as the rp-address and you don't show the config for e1.
That may be an issue.
I'm certainly NOT a multicast expert but here's my take...if
anyone in the group can clarify or correct anything please do. But I
believe when you configure an interface with ip igmp join-group G you
have congigred the router to act as a receiver. So the term Jeff used
"hosting" is, well at least to me, misleading. You're not "hosting" the
groups you're sort of acting like a receiver for traffic destined for
the group.
When you issue a ping to a multicast group address you're
sending or "sourcing" packets from yourself to a mutlticast
address--your're now a multicast traffic "source"! When you ping from
r2 or r3 to 224.1.1.1 you should get a response from r1--do it several
times, it took 5-9 pings for me until it worked. Also, if you issue a
show ip mroute afterward you'll see a (*,G) entry for 224.1.1.1 and an
(S,G) entry--maybe 2 entries on r2 one for each interface configured for
ip pim--with the local router as the source. One will probably have a
null Outgoing Inteface List (OIL) and the other should have the
interface through which the rp is reachable...maybe?
Assuming the rp is reachable this set-up will work. I recently
tested a very similar scenario. But, if the rp-address is not reachable
by your r3 it can't register with the rp as a source for the group so I
don't think the ping will work. The first packet(s) are encapsulated in
a PIM register and then unicast to the rp. If the rp has receivers
(which it does in your case..e0 of r1) it then joins the shortest-path
tree and then kills the unicast registers from continuing with a
register stop.
Oh yea, you probably should make sure both r2 and r3 have the
two dense-mode groups 224.0.1.39 and 224.0.1.40 in the mroute table. I'm
thinking you'll have both a (*,G) and a S,G) entry for both groups on
both routers. If not, even if the rp was reachable, the router won't
know the ip address of the rp yet...clearly a problem
I rambled a bit too much but I hope this helps. As always,
don't take any advice from this list as truth without testing it
yourself first!
Aloha, Frank
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
Tony H.
Sent: Friday, August 02, 2002 7:29 AM
To: Szeto Jeff; ccielab@groupstudy.com
Subject: Re: Multicast question
Hi
Design wise R2 should be hosting the 2 groups and you
will find all things pingable
However with r1 hosting the 2 groups, r2 will prune
back the multicast route to r3. This is normal
behavior for Sparse mode as to decrease multicast
traffic. If you were to add some clients off r3, then
r2 will go from prune to active (join) and r3 should
be able to ping r1
--- Szeto Jeff <jytszeto@hotmail.com> wrote:
> Hi,
>
> I am practising multicast and cannot do a simple
> config.
>
>
>
--e0--R1--s0-------HDLC------s1-R2-S0------HDLC-----s1-R3
>
> R1 ethernet joined two groups:224.1.1.1 and
> 224.2.1.1
> I use sparse-dense-mode, auto-rp
> The problem is R3 cannot ping both multicast groups
> while
> R2 does.
>
> Please help, the related config is at the end,
>
> Thank you.
>
> Regards,
>
> Jeff
>
> R1
> ip multicast-routing
>
> interface Ethernet0
> ip address 172.16.20.2 255.255.255.0
> ip pim sparse-dense-mode
ip igmp join-group 224.1.1.1
ip igmp join-group 224.2.1.1
>
> interface Serial0
> ip address 172.16.12.2 255.255.255.0
> ip pim sparse-dense-mode
> ip pim send-rp-announce Ethernet1 scope 15
> ip pim send-rp-discovery scope 15
>
>
> R2
> ip multicast-routing
> interface Serial0
> ip address 172.16.14.1 255.255.255.0
> ip pim sparse-dense-mode
>
> interface Serial1
> ip address 172.16.12.1 255.255.255.0
> ip pim sparse-dense-mode
>
> R3
>
> ip multicast-routing
> interface Serial1
> ip address 172.16.14.4 255.255.255.0
> no ip directed-broadcast
> ip pim sparse-dense-mode
>
>
This archive was generated by hypermail 2.1.4 : Sat Sep 07 2002 - 19:48:15 GMT-3