[RAM] Re: LISP-01 I-D - UDP vs. IP-in-IP encapsulation

[Robin Whittle <rw at firstpr.com.au>]

Dino wrote:

> That is correct. We are using UDP encapsulation with a LISP
> header for the following reasons:
>
>  o Get through firewalls
>  o ITR hashes on inner header to produce a source port LAG
>    router can hash on
>  o Can carry nonce for ETR anti-spoofing
>  o Can carry Locator reachability bits

With Ivip, there is no TE functionality in the ITR, no
reachability detection of the ETR and no signalling one way or the
other between ITR and ETR.  All that happens is the packet is
tunneled to an IP address which the end-user selects - presumably
the address of an ETR.  So Ivip doesn't need the last three things.

I don't know enough about firewalls in all the various situations
to know what problems IP-in-IP would suffer.

What sort of firewalls would an ETR be located behind?

If an edge network wanted to put an ETR behind a firewall, maybe
they should configure their firewall to handle a major
architectural aspect of getting packets to the ETR - the simplest
and most standardised of which is IP-in-IP, as far as I know.

I am not arguing against UDP encapsulation, except that it is
longer, since it requires a UDP header too.  Ivip could easily use
UDP instead of IP-in-IP.


By the way, links to the current LISP I-Ds are at:

  http://www.firstpr.com.au/ip/sram-ip-forwarding/#LISP

The RRG wiki:

  http://www3.tools.ietf.org/group/irtf/trac/wiki/RoutingResearchGroup

doesn't yet point to all three LISP I-Ds.

 - Robin


_______________________________________________
RAM mailing list
RAM at iab.org
https://www1.ietf.org/mailman/listinfo/ram