It's because the MPLS and GRE are done by the sup and it needs
two passes to do the VRF lookup pre and post encapsulation from
what I remember.
Rodney
On Wed, Jun 25, 2008 at 08:14:33AM +0100, Timothy Arnold wrote:
>
> >Does the addition of the global command 'mls mpls tunnel-recir' make any
> >
>Does the addition of the global command 'mls mpls tunnel-recir' make any
>difference? I've seen some weirdness before with GRE tunnels and MPLS
>without this command.
And my problems magically go away. As soon as I enabled the command, I was able
to see the rest of the VRF
Thanks Matthew!
In article <[EMAIL PROTECTED]>,
<[EMAIL PROTECTED]> wrote:
> Message: 2
> Date: Tue, 24 Jun 2008 15:51:31 +0100
> From: Timothy Arnold <[EMAIL PROTECTED]>
> Subject: Re: [c-nsp] GRE/MPLS
> To: Ozgur Guler <[EMAIL PROTECTED]>, "cisco-nsp@puck.nether
From: Ozgur Guler [mailto:[EMAIL PROTECTED]
Sent: 24 June 2008 14:50
To: cisco-nsp@puck.nether.net; Timothy Arnold
Subject: Re: [c-nsp] GRE/MPLS
Hi Tim,
Is it possible to test it on the 7600 after enabling debug mpls packet?
Cheers
-Ozgur
--- On Tue, 24/6/08, Timothy Arnold <[EMAIL PROTEC
Hi Tim,
Is it possible to test it on the 7600 after enabling debug mpls packet?
Cheers
-Ozgur
--- On Tue, 24/6/08, Timothy Arnold <[EMAIL PROTECTED]> wrote:
From: Timothy Arnold <[EMAIL PROTECTED]>
Subject: Re: [c-nsp] GRE/MPLS
To: "cisco-nsp@puck.nether.net"
Date:
>This definitely sounds label-related - 'regular' routing works, but
>anything requiring label imposition doesn't.
The plot thickens.
I've introduced a 7301 as the tunnel concentrator so it looks like this
CPE---PE(7301)---PE1---PE2---FIREWALL
(appreciate that PE1 is now a 'P' router as it is
Hi Aaron,
>Try adding 'tunnel vrf XXX' as well.
>Everything I have read mentions that this command defines which routing
>table the router will use to reach the 'tunnel destination', but I had
>similar symptoms to you (one way traffic) and this fixed me right up...
The GRE tunnel should use the g
Hi Tim,
>Is the FE interface with the tunnel destination *just* running vanilla IP,
>or also label-switching?
On the CPE, it is vanilla IP. The CPE is not aware of any label switching. The
PE tunnel is using a loopback interface (which we use for management) and
doesn't have label switching tur
Hi Tim,
Sounds a bit like what I hit about a week ago,
> interface Tunnel1
> ip vrf forwarding XXX
> ip address 172.31.255.45 255.255.255.252
> ip mtu 1460
> tunnel source Loopback0
> tunnel destination xxx.xxx.xxx.xxx
> end
Try adding 'tunnel vrf XXX' as well.
Everything I have read mention
Hi Tim,
On Tue, June 24, 2008 10:30 am, Timothy Arnold wrote:
> PE1/PE2 are 7600s with RSP-720s running 12.2(33)SRC. CPEs are 1841s with
> 12.4(19b) and the advanced IP services feature pack.
>
> The CPE has a static default towards PE1 via the GRE tunnel. It has
> another route via the FE interf
Hi Peter,
Thanks for the response.
> What platform/software is the CPE and the PEs? Can you do a trace from
> the CPE and see at least PE1? Does the CPE use a static default towards
> PE1?
PE1/PE2 are 7600s with RSP-720s running 12.2(33)SRC. CPEs are 1841s with
12.4(19b) and the advanced IP ser
On Tue, 2008-06-24 at 09:21 +0100, Timothy Arnold wrote:
> Hi,
>
> I've got to terminate a remote site in to a customer's VRF using a GRE
> tunnel. The tunnel comes up and I can ping across the tunnel and I can
> also reach other hosts in the same VRF on the PE router that the
> tunnel terminat
Hi,
I've got to terminate a remote site in to a customer's VRF using a GRE tunnel.
The tunnel comes up and I can ping across the tunnel and I can also reach other
hosts in the same VRF on the PE router that the tunnel terminates. However I
cannot reach any hosts on other PE routers. For example
13 matches
Mail list logo
