From: Roman Rodichev (roman@iementor.com)
Date: Thu Jun 05 2008 - 00:27:07 ART
No, I got what you were saying, but for a specific VRF do you have RP on
both sides (PE1 *and* PE2)? You can only have PE1 *or* PE2 as an RP.
Roman Rodichev
5xCCIE #7927 (R&S, Security, Voice, Storage, Service Provider)
Instructor, Content Developer. ieMentor Corporation
http://www.iementor.com <http://www.iementor.com/>
Y!M: roman7927
From: Dale Kling [mailto:dalek77@gmail.com]
Sent: Wednesday, June 04, 2008 10:24 PM
To: Roman Rodichev
Cc: Peter Svidler; dara tomar; ccielab@groupstudy.com
Subject: Re: Multicast VPN with MDT
I made PE1's subinterface on that particular VRF the RP address. I think I
mislead you a bit earlier, but for testing multi-vrf simultaneously I was
just going to use the VRFs corresponding subinterface to the CE router as
the RP address for each VRF. Effectively only one RP for each VRF. I don't
really like like placing the RP on one side of the WAN like that, but I
think it will be fine as long as I setup the control plane before I start
streaming huge amounts of multicast end-to-end.
Is there and industry standard as far as RP placement when doing VRFs across
a big CORE like this?
thanks for all the info thus far Roman, you da man. :)
Dale
On Wed, Jun 4, 2008 at 11:17 PM, Roman Rodichev <roman@iementor.com> wrote:
Are you saying you have configured more than one RP for the customer
multicast network? You mentioned "made the PE VRF subinterfaces the RP
addresses". You can only one have one static RP on the customer network.
Regarding SSM, I would need to see your complete configurations to see why
SSM didn't work.
Roman Rodichev
5xCCIE #7927 (R&S, Security, Voice, Storage, Service Provider)
Instructor, Content Developer. ieMentor Corporation
http://www.iementor.com
Y!M: roman7927
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of Dale
Kling
Sent: Wednesday, June 04, 2008 10:14 PM
To: Roman Rodichev
Cc: Peter Svidler; dara tomar; ccielab@groupstudy.com
Subject: Re: Multicast VPN with MDT
Are you saying I only need SSM on my PEs and no the Ps? Everything I Read
had SSM configured throughout the CORE, so unfortunately I didn't know any
better. This is just a test bed for now thank goodness, but this is what I
have.
SW1---CE1---SW2---PE1---P1---P2---PE2---SW3---CE2----SW4
SW=3750s
CE=3845
PE and P= 7604s
I have pim Sparse configured on each VRF on the CEs to the PEs. I did
static RPs for now and made the PE VRF subinterfaces the RP addresses. From
PE1 through the core to PE2 I configured "ip pim ssm". This didn't work
until I made P1 an RP and the rest of the PEs point to it. Something was
wrong with my SSM, but the guides I had didn't explain much of anything.
They just had ip pim ssm configured on the PEs and Ps and said SSM was the
easiest to implement in the CORE. hehe
regards,,
Dale
On Wed, Jun 4, 2008 at 10:59 PM, Roman Rodichev <roman@iementor.com> wrote:
> PIM SSM has to be configured on all edge routers, and doesn't have to be
> configured on core routers. SSM has to be configured on the routers
> connected to SSM-group receivers, other routers don't care about SSM, they
> just participate in normal PIM JOIN tree building regardless if it's
SPARSE
> or SSM.
>
> Roman Rodichev
> 5xCCIE #7927 (R&S, Security, Voice, Storage, Service Provider)
> Instructor, Content Developer. ieMentor Corporation
> http://www.iementor.com
> Y!M: roman7927
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> Dale
> Kling
> Sent: Wednesday, June 04, 2008 9:34 PM
> To: Peter Svidler
> Cc: dara tomar; ccielab@groupstudy.com
> Subject: Re: Multicast VPN with MDT
>
> Not to steal your thread, but it's funny you mention all this. I just
> configured multicast end-to-end on my Multi-VRF CE setup. First time ever
> and what a nightmare, took me about a week. I downloaded a few Cisco
Docs,
> but they seem old. One specifically on multi-vrf multicast.
>
> For some reason I couldn't get SSM working in my CORE, so I switched to
PIM
> Sparse and everything worked beautifully. I found this puzzling
> considering
> all I thought you had to do was configure ip pim ssm on all the Core
> routers
> with PIM sparse on the interconnections of course. According to the DOCs
I
> read I thought that was about all you had to do. My mdt default was
> 232.1.1.1.
>
> I wouldn't mind getting my hand on some better DOCs or book
recommendations
> out there if there are any. I've checked out some Cisco Press Books and
> man
> some of them bad boys are way overkill and a bit too technical at this
> point
> for me. It would be nice to just get some nice explanations on the
> mechanics of how this stuff is working in the background.
>
> In any case, I feel your pain and look forward to doing SP after my R&S.
>
> regards,
>
> Dale
>
>
> On Wed, Jun 4, 2008 at 8:57 AM, Peter Svidler <doubleccie@yahoo.com>
> wrote:
>
> > that is exactly what i was expecting ..however i got an error message
> says
> > that MDT is using interface x for IP address x.x.x.x which is
> non-loopback
> > interface.
> >
> > the BGP is using the loopback in both IPV4 and VPNv4 , and PIM is
enabled
> > on the loopback
> >
> > I have seen similar post to similar problem in the archieve ..however
> have
> > not seen any answer to it
> >
> >
> >
> > --- On Wed, 6/4/08, dara tomar <wish2ie@gmail.com> wrote:
> >
> > From: dara tomar <wish2ie@gmail.com>
> > Subject: Re: Multicast VPN with MDT
> > To: "Peter Svidler" <doubleccie@yahoo.com>
> > Date: Wednesday, June 4, 2008, 8:39 AM
> >
> >
> > Hi Peter,
> >
> > It should use the same loopback address as the BGP with PIM enabled on
> > it!!!
> >
> > What's the issue, that you are facing, can you elaborate please !!!
> >
> > Regards,
> > Dara
> >
> >
> > On Wed, Jun 4, 2008 at 5:00 PM, Peter Svidler <doubleccie@yahoo.com>
> > wrote:
> >
> > Folks
> > I have a scenario where I am trying to build Multicast VPN using MDT
> > between PE's .
> > one of my routers , strangely , is not using the loopback address to
> > initiate the tunnel (using MDT as explained in documentation should
> usually
> > use the BGP peer address , which in my case is the loopback to intiate
> the
> > tunnel )
> > I wonder if anyone came across similar problem and explain why it
> happens?
> >
> > thanks
> >
> >
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
> >
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
>
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Tue Jul 01 2008 - 06:23:20 ART