From: Tim Fletcher (tim@fletchmail.net)
Date: Wed Dec 18 2002 - 13:50:42 GMT-3
Adarsh,
Unless I missed something, you are correct. As far as why the traffic
between r1 & r2 should increase, I'm confused also.
Anybody else have any ideas?
-Tim Fletcher
At 09:59 PM 12/18/2002 +0530, Adarsh Singh wrote:
>Hi Tim ,
>
>Thanks for the reply , I am still very confused on the same , let me tell
>you my doubt.
>
>
>MServer--E0--(router1)--se0--------se0(router2)---se1(multiples gre
>tunnels on serial1)----Service provider Cloud--------Se0(gre
>tunnel)(Router3)--e0---Receiver.
>
>
>
>Now I making loopback interface in router 2 as RP(rendezvous point).
>Mserver sends udp multicast specifically on 239.255.255.255.
>
>So we have STP( shortest transfer path) between router 1 & router 2.
>
>Router 2 has multiple gre tunnels on serial 1 connected to Service
>provider cloud, which terminates at router 3 where we have the receiver .
>
>All the E0, Se0 on router 1 & se0,Se1 on router 2 is running ip
>pim sparse-mode as well as Se0 & E0 of the router 3.
>
>So router 2 should replicate the 239.255.255.255 multicast receiving on se
>0 & Send it through various gre tunnels.
>
>Now the traffic between Router 1 & router 2 should not increase it should
>remain constant , but the traffic between se1(router2) & Serial 1 of
>Service provider end would increase,as you said the encapsulation of the
>packets is taking place at router 2.
>
>Am I correct in stating the above statement.
>
>Pls do let me know.
>
>
>Rgds
>Adarsh Singh
>---- Original Message -----
>From: "Tim Fletcher" <<mailto:tim@fletchmail.net>tim@fletchmail.net>
>To: "Adarsh Singh"
><<mailto:adarshs@hclinfinet.net>adarshs@hclinfinet.net>; "Joe Chang"
><<mailto:changjoe@earthlink.net>changjoe@earthlink.net>;
><<mailto:ccielab@groupstudy.com>ccielab@groupstudy.com>
>Sent: Wednesday, December 18, 2002 8:59 PM
>Subject: Re: Multicast ISSUE
>
> > Adash,
> >
> > In order to get any bandwidth reduction benefit from multicast, you
> have to
> > have a shared media. Tunnels are (virtual) point to point media. The
> router
> > will do the replication, but it will then have to encapsulate each packet
> > in each tunnel and send them separately. There is no way to send a packet
> > through tunnel 1 and have it arrive at destination 2.
> >
> > -Tim Fletcher
> >
> > At 10:45 AM 12/18/2002 +0530, Adarsh Singh wrote:
> > >Hi Joe ,
> > >
> > >Yes the scenario is as you have described the source is multicast
> server on
> > >LAN & the receivers are through tunnel from router 1, is there a way
> I can
> > >decrease the traffic at the serial interface when both receivers are
> using
> > >the multicast traffic , I am expecting router 1 to do the replication
> of the
> > >traffic going tunnels, which is not happening , since the traffic on the
> > >serial keeps increasing as I increase the number tunnels, is there a way
> > >around to keep the traffic between the Router to Router1 constant to the
> > >multicast traffic , inspite of adding number of tunnels.
> > >
> > >
> > >Rgds
> > >Adarsh
> > >----- Original Message -----
> > >From: "Joe Chang" <<mailto:changjoe@earthlink.net>changjoe@earthlink.net>
> > >To: <<mailto:ccielab@groupstudy.com>ccielab@groupstudy.com>
> > >Sent: Tuesday, December 17, 2002 11:34 PM
> > >Subject: Re: Multicast ISSUE
> > >
> > >
> > > > It seems that you have enabled pim on the physical interfaces
> between all
> > > > the routers, as well as between the tunnel endpoints (assuming the
> tunnels
> > > > for R2 and R3 end at the multicast server.) When the network first
> starts
> > >up
> > > > the leaf routers will have to choose whether to use its tunnel or
> physical
> > > > interface as its RPF interface for group 239.255.255.255. Since the
> tunnel
> > > > is only one hop back to the source, the prune will be requested for the
> > > > physical interface. The server router will then prune its serial
> interface
> > > > from its oil list, and have to send out one multicast packet for each
> > > > tunnel, thereby doubling the traffic.
> > > >
> > > > ----- Original Message -----
> > > > From: "Adarsh Singh"
> <<mailto:adarshs@hclinfinet.net>adarshs@hclinfinet.net>
> > > > To: "Don" <<mailto:seadon@attbi.com>seadon@attbi.com>; "Sean
> Garrett" <<mailto:SGarrett@cnitech.com>SGarrett@cnitech.com>
> > > > Cc: <<mailto:ccielab@groupstudy.com>ccielab@groupstudy.com>
> > > > Sent: Tuesday, December 17, 2002 1:40 AM
> > > > Subject: Multicast ISSUE
> > > >
> > > >
> > > > > Hi Don,
> > > > >
> > > > > I have this typical problem with multicast :
> > > > >
> > > > > Multicast Server(router)----LL(64Kbps)----Router
> > >1-----tunne1 ----R2(each
> > > > > tunnel traffic is 20 Kbps)
> > > > >
> > > > > ------tunnel2 ---R3
> > > > >
> > > > >
> > > > > Now as the number of tunnels(gre) increases at the router 1 , the
> > >traffic
> > > > > increases propotinally for the 64 Kbps leased line .
> > > > > So the mulicast is not taking place in real sense .
> > > > >
> > > > >
> > > > >
> > > > > The commands used at tunnel are as follows :
> > > > >
> > > > > interface Tunnel2
> > > > > ip address 172.168.12.1 255.255.255.252
> > > > > ip directed-broadcast
> > > > > ip pim sparse-dense-mode
> > > > > ip igmp join-group 239.255.255.255
> > > > > load-interval 30
> > > > > tunnel source 10.65.12.233
> > > > > tunnel destination 10.74.5.9
> > > > >
> > > > > interface FastEthernet0
> > > > > description connected to EthernetLAN
> > > > > ip address 192.168.10.85 255.255.255.0 secondary
> > > > > ip directed-broadcast
> > > > > ip pim sparse-dense-mode
> > > > > ip igmp join-group 239.255.255.255
> > > > > load-interval 30
> > > > > speed auto
> > > > > half-duplex
> > > > >
> > > > >
> > > > > interface Serial1
> > > > > description connected to infinet for VPN
> > > > > ip address 10.65.7.82 255.255.255.252
> > > > > ip accounting output-packets
> > > > > ip pim sparse-dense-mode
> > > > > ip igmp join-group 239.255.255.255
> > > > > load-interval 30
> > > > >
> > > > >
> > > > > Do let me know if someone has ideas of where i an making a
> mistake.
> > > > >
> > > > >
> > > > > Rgds
> > > > > Adarsh Singh
> > > > > .
> > > > .
> > >.
> > ><mailto:majordomo@groupstudy.com>majordomo@groupstudy.com with the
> body containing:
> > >unsubscribe ccielab
.
This archive was generated by hypermail 2.1.4 : Fri Jan 17 2003 - 17:21:48 GMT-3