[j-nsp] v6-in-v4 tunnels without tunnel PIC?

[Paul Goyette]

Use of RE-based tunnel interfaces is supported only for the
GMPLS control channel;  all other uses of host-based tunnels
is unsupported.

Alse please remember that packets that arrive via a PFE 
interface are never forwarded out of an RE interface, and
vice versa.  So, for example, a packet that arrives on a
ipip interface on the RE will not get forwarded to a PFE
interface.  This is a logical corrollary of packets not
being forwarded between the PFE interfaces and fxp0....

Yup, you need a tunnel PIC!

-----Original Message-----
From: juniper-nsp-bounces at puck.nether.net
[mailto:juniper-nsp-bounces at puck.nether.net]On Behalf Of Joe Abley
Sent: Thursday, September 23, 2004 6:17 PM
To: Phil Rosenthal
Cc: juniper-nsp at puck.nether.net
Subject: Re: [j-nsp] v6-in-v4 tunnels without tunnel PIC?



On 23 Sep 2004, at 19:37, Phil Rosenthal wrote:

> interfaces {
>     ipip {
>         unit 101 {
>             description "IPv6 Tunnel to (blah)";
>             tunnel {
>                 source 1.2.3.4;
>                 destination 5.6.7.8;
>             }
>             family inet6 {
>                 address 2001:d00d:b00b::babe/126;
>             }
>         }
>     }
> }
>
> Requires no tunnel services card -- I have no idea of the performance 
> characteristics of this, I have been unable to accurately test.

I tried this out.

It looks very much to me like this works in as far as you can reach the 
RE at the other end of the tunnel, and it responds to pings as you'd 
expect. I have not been able to convince a packet sent down the tunnel 
towards an "ipip" termination to continue out a real interface, though.

It seems likely/possible that this is an architectural problem, and the 
solution to it is "buy a tunnel PIC".


Joe

_______________________________________________
juniper-nsp mailing list juniper-nsp at puck.nether.net
http://puck.nether.net/mailman/listinfo/juniper-nsp