From: John Elias (jelias_@xxxxxxxxxxx)
Date: Fri Jan 04 2002 - 16:42:05 GMT-3
Guys,
- In a tokenring enviorment I believe you need the interface command, "ip
multicast use-functional" on the router.
- The interface command, "ip pim nbma-mode", also needs to be there for a
NBMA setup. "Configuration is the same as in a regular multicasting
environment except with a few extra commands. The ip pim nbma-mode is
used in a Frame-Relay environment.", this is from Beau Williamson book.
John E.
CCIE #8150
>From: "D. J. Jones" <meganac@attbi.com>
>Reply-To: "D. J. Jones" <meganac@attbi.com>
>To: "Aiman A. M." <drpower@hotmail.com>, <Ron.Kirby@getronics.com>,
><ccielab@groupstudy.com>
>Subject: Re: Multicast over NBMA
>Date: Fri, 4 Jan 2002 08:43:34 -0800
>
>I believe there is a performance impact of token ring mapping of multicast
>addresses. According to Beau Williamson (Developing IP Multicast Networks
>Volume 1, pp 33-34) "Many Token Ring NICs do not have the capabiity to be
>programmed to interrupt the CPU when an arbitrary multicast MAC address is
>received. Therefore, the only alternative is to use the all-ones
>(0xffff.ffff.ffff)
>broadcast MAC address or a functional address for IP multicast to MAC
>address
>mapping as hown in figure 2.7." Further, on page 34 he states "This means
>all
>multicast traffic on the ring will cause a Token Ring node's CPU to be
>interrupted
>for every multicast packet on the ring!"
>
>"In addition to causing this CPU performance impact, the all-into-one
>mapping scheme
>virtually precludes Token Ring switches from constraining multicast
>flooding
>at the
>MAC layer. As a result, Token Ring switches, which operate solely at layer
>2, are
>powerless to prevent any Token Ring node that joins a multicast group and
>begins
>receiving multicast traffic from afffecting every other node on the ring."
>
>Bottom line is that you may wish to pursue a course of action that involves
>a switched
>ethernet environment instead of token ring.
>----- Original Message -----
>From: "Aiman A. M." <drpower@hotmail.com>
>To: <Ron.Kirby@getronics.com>; <ccielab@groupstudy.com>
>Sent: Friday, January 04, 2002 3:58 PM
>Subject: Multicast over NBMA
>
>
> > hi all,
> > how to do multicasting in token ring enviroment ?, thanx.
> >
> > Aiman
> >
> >
> > >From: "Kirby, Ron" <Ron.Kirby@getronics.com>
> > >Reply-To: "Kirby, Ron" <Ron.Kirby@getronics.com>
> > >To: ccielab@groupstudy.com
> > >Subject: Multicast over NBMA
> > >Date: Fri, 4 Jan 2002 10:44:23 -0500
> > >
> > >All,
> > >
> > >I am working with multicasting over a frame cloud. R1 has a multipoint
> > >subinterface connected to R3 and R4 which are using physical interfaces
> > >(basic hub and spoke design). I am using EIGRP and have split horizon
> > >turned
> > >off at the hub (R1). The routing table is fine and everything can be
> > >pinged.
> > >
> > >Now, Pim is running on all routers and dense mode on all interfaces. I
> > >have
> > >pim neighbor relationships between R1 and R3 and R1 and R4. Here's the
> > >problem. R1 (the hub) has an mroute entry for the multicast source off
>of
> > >R3 as shown below:
> > >
> > >**********************************************
> > >r1#sh ip mroute 234.5.6.7
> > >IP Multicast Routing Table
> > >Flags: D - Dense, S - Sparse, C - Connected, L - Local, P - Pruned
> > > R - RP-bit set, F - Register flag, T - SPT-bit set, J - Join
>SPT
> > > M - MSDP created entry, X - Proxy Join Timer Running
> > > A - Candidate for MSDP Advertisement
> > >Outgoing interface flags: H - Hardware switched
> > >Timers: Uptime/Expires
> > >Interface state: Interface, Next-Hop or VCD, State/Mode
> > >
> > >(*, 234.5.6.7), 00:12:31/00:02:59, RP 0.0.0.0, flags: D
> > > Incoming interface: Null, RPF nbr 0.0.0.0
> > > Outgoing interface list:
> > > Ethernet0/0, Forward/Dense, 00:12:09/00:00:00
> > > Serial0/0.2, Forward/Dense, 00:12:31/00:00:00
> > > Serial0/0.1, Forward/Dense, 00:12:31/00:00:00
> > >
> > >(192.168.30.2, 234.5.6.7), 00:02:46/00:00:24, flags: PT
> > > Incoming interface: Serial0/0.2, RPF nbr 131.5.20.3
> > > Outgoing interface list:
> > > Serial0/0.1, Prune/Dense, 00:02:46/00:00:19
> > > Ethernet0/0, Prune/Dense, 00:02:46/00:00:19
> > >***************************************************
> > >
> > >
> > >As you can see, S0/0.2 is not in the outgoing interface list, and I
>figure
> > >it is because the router does it's RPF check and finds that the
>upstream
> > >interface is S0/0.2 and therefore S0/0.2 is not a candidate for
>flooding
> > >the
> > >multicast data. So how do I get the multicast traffic to a host on R4?
>I
> > >believe my possibilities are a tunnel between R1 and R3 and R1 and R4
>for
> > >multicast purposes, or moving to frame-relay P-to-P subinterfaces. I
>have
> > >also worked with the NBMA mode command, but it doesn't look as if it is
> > >designed for this. Is there a command line solution available?
> > >
> > >Thanks
> > >Ron
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:56:16 GMT-3