[Xorp-users] inter as multicast

Pavlin Radoslavov pavlin@icir.org
Fri, 24 Jun 2005 12:57:16 -0700


> I am trying to configure one network with 2 Autonomous Systems like it shows
> 
> 
>  
> 
>                             AS1                              |        |
> AS2
> 
> Host1 --------- cisco1 --------- xorp1 ---|------|---- xorp2 -----------
> cisco2 ---------- host2 
> 
>                                                                   |        |
> 
>  
> 
> In cisco1 I have configured RIP, MSDP and PIM SM
> 
> In cisco2 I have configured RIP, MSDP and PIM SM
> 
> In xorp1 I have configured RIP, MBGP and PIM SM
> 
> In xorp2 I have configured RIP, MBGP and PIM SM
> 
>  
> 
> The configuration files are in attachment.
> 
>  
> 
> When I start a multicast session with host1 I see MSDP SA packets between
> cisco1 and cisco2. The multicast session is to 224.30.30.30.
> 
> When host2 joins to the session I see cisco2 sending a join (S, G) message

Do you really mean (S,G) Join message or is it (*,G) Join?
The (S,G) Join message should have the host1 address as S, while
(*,G) join message would have the RP address instead (either
192.10.10.10 or 192.20.10.10) and the RPT and W flag set. Please
double-check by running tcpdump between xorp1 and xorp2.

If the message is (*,G) Join, then xorp1 will silently drop
it because of the RP address mismatch (the RP inside the (*,G) Join
message doesn't match the RP address for that group according to
xorp1). E.g., see Section 4.5.2 of the lastest PIM-SM I-D spec
(draft-ietf-pim-sm-v2-new-11.txt):

====
When a router receives a Join(*,G) or Prune(*,G) it must first check to
see whether the RP in the message matches RP(G) (the router's idea of
who the RP is).  If the RP in the message does not match RP(G) the
Join(*,G) or Prune(*,G) should be silently dropped.
====


If the message is (S,G) Join as you say, then indeed xorp1 must
contain a matching entry for host1 in the MRIB, otherwise the (S,G)
Join won't be propagated.
If there was no matching MRIB entry, typically XORP should print a
warning log message "no upstream PIM neighbors" or something like
this.

In your case the route to host1 should come from RIP, so you can
use the following commands to locate the missing 192.10.20.0/24
route on xorp1:

show route table ipv4 unicast rip
show route table ipv4 unicast final
show route table ipv4 multicast fib2mrib
show route table ipv4 multicast final
show pim mrib

Pavlin


> to xorp2, a see the same between xorp2 and xorp1 but I don't see any join
> message between xorp1 and cisco1, resulting to fail the association of host2
> to the multicast session.
> 
> I have seen the multicast table of xorp1 and I don't   find any entry to the
> net where host1 is present. I think this is why xorp1 don't send the join
> message to cisco1.
> 
>  
> 
> My question is am I doing something wrong in my configuration???