From: Tony H. (aamercado31@xxxxxxxxx)
Date: Wed Jun 26 2002 - 21:41:59 GMT-3
Hi
Scenario #1 (without auto rp)
sparse-dense mode without rp assumes dense mode
whereby you should be able to ping everywhere that is
multicast enabled due to the multicast flooding
involved
Scenario #2 (with auto rp)
sparse-dense mode with rp limits the flooding and
shifts you to "spare-mode". At the same time, you will
now see the 2 well-known groups 224.0.1.39 and
224.0.1.40. RPs send their announcements using
224.0.1.39, while the RP mapping agent listens to the
announced packets from the RPs, then sends RP-to-group
mappings in a discovery message that is sent to
224.0.1.40. Because you are in "sparse mode" now, it
is advisable to configure "ip pim nbma-mode" on each
router. At any point, R4 will not respond to a ping
because it has been "pruned" back which is the nature
of sparse mode.
According the the mroute table:
> > (*, 224.0.1.39), 00:19:43/00:02:59, RP 0.0.0.0,
> flags: DJC
> > Incoming interface: Null, RPF nbr 0.0.0.0
> > Outgoing interface list:
> > Serial0, Forward/Sparse-Dense,
> 00:19:43/00:00:00
> > Serial1, Forward/Sparse-Dense,
> 00:19:43/00:00:00
Since R1 is not the RP, the RP is 0.0.0.0 and we can
ignore this entry in the mroute table. Basically,
since we are announcing (.39) both outgoing interface
is sending the .39 which is default in sparse mode.
You should see this on all sparse routers.
> > (2.2.2.2, 224.0.1.39), 00:00:43/00:02:16, flags:
> PCT
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list:
> > Serial1, Prune/Sparse-Dense, 00:00:43/00:02:22
The ip add of the RP (2.2.2.2) it sent to R1 with a
reverse path - RPF nbr 1.1.2.2 back to R1 via the
incoming int s0. NOTICE that the outgoing interface S1
is prune back which is the nature of a sparse mode
which explains your pinging problem. There is no PC
host off R4 trying to join so this link is pruned.
Basically R1 will not resend the RP ip add to R4
> > (*, 224.0.1.40), 00:20:25/00:00:00, RP 0.0.0.0,
> flags: DJCL
> > Incoming interface: Null, RPF nbr 0.0.0.0
> > Outgoing interface list:
> > Serial0, Forward/Sparse-Dense,
> 00:20:25/00:00:00
> > Serial1, Forward/Sparse-Dense,
> 00:20:26/00:00:00
Again, Since R1 is not the RP, the RP is 0.0.0.0 and
we can ignore this entry in the mroute table and same
logic applies to above desc on .39.
> > (2.2.2.2, 224.0.1.40), 00:19:53/00:02:59, flags:
> CLT
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list:
> > Serial1, Forward/Sparse-Dense,
> 00:19:53/00:00:00
Here .40 is the mapping agent and sees the rp on
2.2.2.2 with a next hop to 1.1.2.2. NOTICE THAT IT IS
IN FORWARDING MODE just in case there is a pc host of
R4 which wants to recieve traffic
> > (*, 224.0.7.7), 00:00:11/00:02:59, RP 2.2.2.2,
> flags: SP
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list: Null
Finally, 224.0.7.7 route shows up, which announce the
RP with a next hop of 1.1.2.2 via s0. Notice that the
outgoing interface is Null out...sorta similar to the
pruning because there is no PC host of R4 wanting to
join.
> > (1.1.4.4, 224.0.7.7), 00:00:11/00:02:48, flags: PT
> > Incoming interface: Serial1, RPF nbr 1.1.4.4
> > Outgoing interface list: Null
In this last entry, R4 is recognized as a multicast
router via s1 with next hop of 1.1.4.4 but the
outgoing interface is Null out due to the advantage of
sparse mode not to flood.
In regards to changing it to HDLC, this implies ptp
rather than nbma. Also with putting the RP on R1, that
is just good design rather than scenario #2 which is
good lab, but bad design..See:
http://www.cisco.com/univercd/cc/td/doc/cisintwk/intsolns/mcst_sol/frm_rlay.htm
This is what I think happens..Any multicast experts
out there..please advise as I wonder if I got this
right?
HTH
--- Joe Higgins <netsat@optonline.net> wrote:
> Changing the interfaces from r4 to r1 from physical
> frame to point-to-point
> frame subinterfaces fixes this problem. I do not
> know why it fixes the
> problem?
>
> Joe Higgins wrote:
>
> > I have the following problem. I have three
> routers r4, r1 and r2. All
> > have ip multicasting-routing running. R4 is
> connected to r1 via frame.
> > serial 1 to serial 1. R1 is connected to r2 via
> frame, serial 0 to
> > serial 0. All affected interfaces are running ip
> pim sparse-dense. R2
> > has ip igmp join on its e0 interface for multicast
> address 224.0.7.7.
> > Ospf is running between al routers
> > R4-s1 <> R1-s1-R1-s0 <> R2-s0
> >
> > Before I enable auto rp on R2 I can ping 224.0.7.7
> from r4. After I
> > enable auto rp on r2 I cannot ping 224.0.7.7 from
> r4. I have configured
> > both announce and discovery on r2. R1 mroute
> shows the following when
> > it fails.
> >
> > (*, 224.0.1.39), 00:19:43/00:02:59, RP 0.0.0.0,
> flags: DJC
> > Incoming interface: Null, RPF nbr 0.0.0.0
> > Outgoing interface list:
> > Serial0, Forward/Sparse-Dense,
> 00:19:43/00:00:00
> > Serial1, Forward/Sparse-Dense,
> 00:19:43/00:00:00
> >
> > (2.2.2.2, 224.0.1.39), 00:00:43/00:02:16, flags:
> PCT
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list:
> > Serial1, Prune/Sparse-Dense, 00:00:43/00:02:22
> >
> > (*, 224.0.1.40), 00:20:25/00:00:00, RP 0.0.0.0,
> flags: DJCL
> > Incoming interface: Null, RPF nbr 0.0.0.0
> > Outgoing interface list:
> > Serial0, Forward/Sparse-Dense,
> 00:20:25/00:00:00
> > Serial1, Forward/Sparse-Dense,
> 00:20:26/00:00:00
> >
> > (2.2.2.2, 224.0.1.40), 00:19:53/00:02:59, flags:
> CLT
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list:
> > Serial1, Forward/Sparse-Dense,
> 00:19:53/00:00:00
> >
> > (*, 224.0.7.7), 00:00:11/00:02:59, RP 2.2.2.2,
> flags: SP
> > Incoming interface: Serial0, RPF nbr 1.1.2.2
> > Outgoing interface list: Null
> >
> > (1.1.4.4, 224.0.7.7), 00:00:11/00:02:48, flags: PT
> > Incoming interface: Serial1, RPF nbr 1.1.4.4
> > Outgoing interface list: Null
> >
> > On both r1 and r4 a show ip pim rp mapp indicates
> >
> > Group(s) 224.0.0.0/4
> > RP 2.2.2.2 (?), v2v1
> > Info source: 2.2.2.2 (?), via Auto-RP
> > Uptime: 00:34:44, expires: 00:01:59
> >
> > R1 gets the ping but fails to send it to an
> outgoing interface. I know
> > that there is a problem wt. multicast and frame
> but what I have read
> > only refers to coming in and out of the same
> physical interface.
> >
> > Additional facts are as follows. If I make the
> connection from r4 to r1
> > hdlc, it works. If I put the rp announce and
> agent on r1, it works. I
> > have spent a couple of days on this and any help
> would be appreciated.
> > I really hurts not having the search facility
> working on this site.
> > Thanks in advance.
> >
>
This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:42 GMT-3