Thoughts on increasing MTUs on the internet

Fri Apr 13 05:20:37 UTC 2007

On (2007-04-13 00:17 +0100), Will Hargrave wrote:

> At LONAP a jumbo frames peering vlan is on the 'to investigate' list. I
> am not sure if there is that much interest though. Another vlan, another
> SVI, another peering session...

Why another? For neighbours that are willing to peer over eg. VLAN MTU 9k,
peer with them only over that VLAN. I don't see much point peering
over both VLANs.
What I remember discussing with unnamed european IXP staff was that
they were worried about loosing 'frame too big' counters. Since of
course then the switch environment would accept bigger frames even
on the 1500 MTU VLAN. And if member misconfigures the small MTU VLAN,
and calls to IXP complaining how IXP is dropping their frames (due 
to sending over 1500bytes) IXP staff can't quickly diagnose the
problem from interface counters. I argued that it's mostly irrelevant,
since IXP staff can ping from IXP 'small mtu VLAN' the customer
they're suspecting sending too large frames, and confirm this
if router replies to a ping over 1500 bytes. But then again, I have
0 operational experience running IXP and it's easy for me to
oversimplify the issue.

> The fabric itself is enabled to 9216 bytes; we have several members
> exchanging L2TP DSL traffic at higher MTUs but this is currently done
> over private (i.e. member addressed) vlans.

This I believe to be biggest gain, tunneling, eg. ability to run IPSec
site-to-site while providing full 1500bytes to LAN.

> There are some other possible IX applications... MPLS springs to mind as
> another network technology which requires at least baby giants; what
> would providers use this for? Handoff of multiprovider l2/l3 VPNs?
> 
> The other technology which sees people deploying jumbos out there is
> storage. Selling storage as well as transit over the IX? It could happen :-)
> 
> -- 
> Will Hargrave             will at lonap.net
> Technical Director
> LONAP Ltd
> 
> 
> 
> 
> 

-- 
  ++ytti