[AusNOG] Sensible MTU For Core Routers

[James Mcintosh]

Given that edge routers only forward traffic rather then generate it (except for maybe BGP) is there any harm in having the MTU higher than 1500?
After all in this situation it's the client and the server that actually create the packet and associated MTU. The routers in between are just forwarding?

 


     On Sunday, 19 April 2015, 11:54, Nathan Brookfield <Nathan.Brookfield at simtronic.com.au> wrote:
   

 Hi James,
For anything internet facing for compatibility really and the fact that all transit providers and most peers on an exchange would be using 1500 byte MTU's we enforce this at all edges.
For customer Ethernet we enable by default at least 1998 unless requested by the end client for Jumbo's or baby Giants, this allows them to do a bit of encapsulation where required, L2TP etc without complaining to us about clamping.
For our own core network so that we can entertain 9216 at Layer 2 if requested, we ensure we provision every back haul link and interface touching these at 9000 Bytes Layer 3.

Kindest Regards,Nathan Brookfield
Chief Executive OfficerSimtronic Technologies Pty Ltd
Web: http://simtronic.com.auPhone: 1300 592 330Fax: (02) 4749 4950
On 19 Apr 2015, at 10:54, James Mcintosh <james.mcintosh at rocketmail.com> wrote:

Hey Noggers,
With router MTU's being configurable up to a whopping 9216 what are people setting for their interface MTU's on internet connected core and PE routers?
Is it is still just the default of 1500? What about core switches? What are you setting there to get get interoperability with various ethernet providers?

-James_______________________________________________
AusNOG mailing list
AusNOG at lists.ausnog.net
http://lists.ausnog.net/mailman/listinfo/ausnog


  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ausnog.net/pipermail/ausnog/attachments/20150420/9bfb0e7d/attachment-0001.html>