It's not as simple as that. What if the 6rd functionality is running on the access concentrator like a CMTS in cable? In a cable deployment, due to the added len from docsis headers, the MTU in cable between the home and the cable access concentrator is actually less than the typical MTU for Ethernet LANs. So do we really have to specify an MTU for the 6rd specification?
Hemant -----Original Message----- From: softwires-boun...@ietf.org [mailto:softwires-boun...@ietf.org] On Behalf Of Brian E Carpenter Sent: Monday, November 09, 2009 1:21 PM To: Ole Troan Cc: softwires@ietf.org Subject: Re: [Softwires] MTU issue for 6RD (and 6to4) On 2009-11-09 14:12, Ole Troan wrote: >> Should we be specifying a default MTU of 1280 for 6RD too? >> >> (The same question applies retroactively to 6to4, of course.) > > this is what the draft says: > > "6rd's scope is limited to a service provider network. If the MTU is > well-managed > such that the IPv4 MTU on the 6rd CE WAN interface is set so that no > fragmentation occurs within the boundary of the SP, then the IPv6 MTU > should be set to the IPv4 MTU minus the size of the encapsulating IPv4 > header (20 bytes)." > > the assumption being that the service provider maintains a consistent > MTU across its network. Sure. If the v4 MTU is 1500, set the 6rd MTU to 1480. But shouldn't we specify a default for shipped products, and shouldn't it be 1280 (as the least bad choice)? > there is more reason for 6to4 to have a default MTU of 1280 though. I agree, and I wish we had specified a default. All this is orthogonal to PMTUD and ICMP translation; we're talking about a default for when PMTUD fails to work at all. Brian _______________________________________________ Softwires mailing list Softwires@ietf.org https://www.ietf.org/mailman/listinfo/softwires _______________________________________________ Softwires mailing list Softwires@ietf.org https://www.ietf.org/mailman/listinfo/softwires
