Re: [RRG] Tunnel fragmentation/reassembly for RRG map-and-encaps architectures

[Tony Li <tli@cisco.com>]

[Multi-reply...]

On Jan 16, 2008, at 8:10 AM, Noel Chiappa wrote:

> My advice to the LISP people was that they didn't need to implement  
> this
> security stuff in any prototype implemenatations they were doing  
> (where
> manpower is limited, and the chief concern is trying to get some  
> real-world
> data on things like delays, unforseen operational issues, etc),  
> because it
> doesn't change the fundamentals of the operation at all; it's just  
> security
> sugar layered over everything.
>
> However, it has all been thought through and worked out, to make  
> sure that
> the basic architecture they are trying out *is* securable.


As you mentioned elsewhere, there is no progress without pain.  One  
of those pain points is that security is now a much more stringent  
requirement today than it was previously, and having proven, deployed  
security mechanisms as part of a proposal would seem to be a Very  
Good Thing, IMHO.  Just having the claim that something is securable  
is only going to hold so much water and will definitely attract  
skeptics.



On Jan 16, 2008, at 8:33 AM, Brian Dickson wrote:
> We don't need to reinvent the wheel, IMHO - just take advantage of  
> the dnssec.


That would be fine by me.  However, DNSsec has been awaiting  
deployment for about a decade now.  I'm not hopeful that this is  
going to happen anytime soon.

More generally, my friends in the security and operations communities  
point out that, in general, approaches with a full blown PKI  
infrastructure are simply too heavyweight to be pragmatically  
deployable.  There are simply too many interdependencies.  Their  
strong suggestions point much more towards pairwise security and/or  
web-of-trust approaches (ala PGP).


On Jan 16, 2008, at 2:11 PM, Brian Dickson wrote:
> There's two distinct things:
> 1) how to *publish* the data;
> 2) how to *serve* the data
>
> Both need to be secure to be trustworthy.


Is this really true?  Does serving the data truly need to be secure  
as long as the data is authentic, accurate, timely, etc.?

If we can avoid securing the mapping transport layer, it would be a  
very big win.

Tony


--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg