Re: [DNSOP] WGLC for draft-ietf-dnsop-tcp-requirements

[Joe Abley <[e-mail]>]

On 19 Apr 2021, at 12:40, Peter van Dijk <[e-mail]> wrote:

> This note on statelessness is good, but I don't think it should be tied to IPv6. Packets get lost in IPv4 too, especially when they are big, and even if such evens trigger a report in the form of an ICMP message, the same lack-of-state problem applies.

In IPv4, datagrams that need to be transmitted over a link with an MTU is too low are fragmented by the router attached to the link, assuming DF=0. There is no signal sent back to the source in that case. In IPv6 that doesn't happen.

In the v4 case a large DNS message (large enough to require fragmentation along the path) can be transmitted without the source having to retain any state. That's not true in v6.

So I think the v4 and v6 cases are different. That's why I attached that comment to the v6 case.

DNS messages can be lost in both v4 and v6 for a variety of other reasons, I agree.

> https://datatracker.ietf.org/doc/draft-ietf-dnsop-avoid-fragmentation/ even proposes setting DONTFRAG socket options, and some servers out there already send IPv4 replies with the DF bit set (the two I can verify immediately are OpenDNS, and whatever is running on the router my provider gave me, but most likely there are others too).

Setting DF=1 does seem like it would avoid the differences I was trying to allude to above, I agree. With DF=1 fragmentation (or not-fragmentation) is just another reason for a packet to get dropped.


Joe