Re: [RRG] A data point on transit MTU size

[Iljitsch van Beijnum <iljitsch@muada.com>]

On 22 sep 2008, at 23:49, Brian E Carpenter wrote:

> We've argued here about whether it's reasonable to assume >1500 byte
> MTU on transit links, when running a LISP-type solution.

> Here's a data point about the real world, as far as Internet exchange
> points at the southern end go:

> http://list.waikato.ac.nz/pipermail/nznog/2008-September/014471.html

Note that they only talk about the capabilities of their switches.

Obviously jumbo capability there is a prerequisite, but this is easily  
met by pretty much all gigabit and faster gear. The next problem,  
which is harder to crack, is finding a way to send larger packets even  
though some routers on the shared infrastructure don't support this.

I've been working for a while now on a draft that would solve this  
through automatic discovery and testing of the jumboframe capability  
between two systems on a subnet two systems can use the max MTU they  
both support between them without being limited by less capable  
systems on the same subnet:

http://www.muada.com/drafts/draft-van-beijnum-multi-mtu-02.txt

However, for the internet exchange case this would probably be much  
easier to solve through static configuration. Something like static  
ARP entries with an MTU attached that can overrule the default MTU.  
Would any of our favorite vendors be interested in implementing  
something like that?

--
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