[j-nsp] lsp ping between JNPR and Cisco

[Harry Reynolds]

Not to stir the pot, but I've always understand that receipt of a ttl 1
packet for the *local* host should accepted. Not supposed to forward a
ttl 1, but OK to receive it.


Anyway, IIRC, we have a sysctl to alter our default:

% sysctl -a | grep -i ip_ttl_zero_reject
net.inet.ip.ip_ttl_zero_reject: 1
% 

Separate sysctl one for v6. Standard disclaimer as to this not being
supported as shell command etc.



IIRC, we failed in tahai phase II logo test for this with v6, but got a
pass with the sysctl modified (for v6)




Regards


 

-----Original Message-----
From: juniper-nsp-bounces at puck.nether.net
[mailto:juniper-nsp-bounces at puck.nether.net] On Behalf Of Marlon Duksa
Sent: Monday, December 08, 2008 2:20 PM
To: Daniel.Hilj at synetrix.co.uk
Cc: juniper-nsp at puck.nether.net; cisco-nsp at puck.nether.net
Subject: Re: [j-nsp] lsp ping between JNPR and Cisco

When we replaced Csco with JNPR box as transit LSR, the PING worked.
With or without 127.0.0.1.
Obviously there is an interop issue between Csco and JNPR, namely Cisco
is decrementing IP TTL as penultimate hop. And we don't know how to
disable this...
Thanks,
Marlon


On Sat, Dec 6, 2008 at 12:20 PM, <Daniel.Hilj at synetrix.co.uk> wrote:

> I take it that you already configured 127.0.0.1 on the loopbacks which

> is required for MPLS ping to work on Junipers?
>
> Regards
> Daniel
>
>
> -----Original Message-----
> From: Marlon Duksa <mduksa at gmail.com>
> Sent: 05 December 2008 22:57
> To: Juniper-Nsp <juniper-nsp at puck.nether.net>; 
> cisco-nsp at puck.nether.net < cisco-nsp at puck.nether.net>
> Subject: [j-nsp] lsp ping between JNPR and Cisco
>
> Our RSVP tunnel endpoints are JNPR boxes (M320) and a transit node is 
> Cisco (7600). When we try to initiate MPLS ping from JNPR to JNPR 
> through Cisco, the mpls ping fails.
> The reason is that JNPR is always setting IP TTL as 1. Since the Cisco

> is a penultimate node, it strips the label, decrement the IP TTL (to 
> 0) and send the packet to JNPR. JNPR discards it since the IP TTL is
0.
>
> Does anyone know if there is any workaround to this?
>
> It looks to me that the only option is to try to set the IP TTL in 
> MPLS ping from ingress JNPR to something > 0. Unfortunately there is 
> no provision that would allow us to do this.
> On the other hand, Cisco won't honor 'no-ttl-decrement' statement on 
> the penultimate if MPLS TTL is greater then the IP TTL (which 
> currently is since JNPR MPLS TTL is set to 255).
>
>
> Thanks,
> Marlon
>
>
_______________________________________________
juniper-nsp mailing list juniper-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp