Re: [tsvwg] UDP options and header-data split (zero copy)

[Tom Herbert <[e-mail]>]

On Sun, Aug 1, 2021, 3:48 PM Joseph Touch <[e-mail]> wrote:

> Hi, Tom,
>
> > On Aug 1, 2021, at 3:39 PM, Tom Herbert <[e-mail]> wrote:
> >
> > There is also RFC9000:
> >
> > "QUIC assumes a minimum IP packet size of at least 1280 bytes.  This
> > is the IPv6 minimum size [IPv6] and is also supported by most modern
> > IPv4 networks.
>
> Hmm. Seems like they completely overlooked IP source fragmentation support
> and 1500B reassembly...
>

No IP fragmentation in QUIC. DF bit is always set.


> > Assuming the minimum IP header size of 40 bytes for
> > IPv6 and 20 bytes for IPv4 and a UDP header size of 8 bytes, this
> > results in a maximum datagram size of 1232 bytes for IPv6 and 1252
> > bytes for IPv4.  Thus, modern IPv4 and all IPv6 network paths are
> > expected to be able to support QUIC."
>
> QUIC over UDP with fragmentation wouldn’t be bothered by these limits at
> all; it could send a 1200B QUIC payload with up to 63KB of additional
> combined IP and UDP headers - not that it should, but it could.
>
> By the time QUIC sees the packet, it would be both IP reassembled (at
> least 1500B) and, with UDP options, UDP reassembled (up to a total of 64K).
>
> > and
> >
> > "This requirement to support a UDP payload of 1200 bytes limits the
> > space available for IPv6 extension headers to 32  bytes or IPv4
> > options to 52 bytes if the path only supports the IPv6 minimum MTU of
> > 1280 bytes.  This affects Initial packets and path validation."
>
> See above.
>
> > If UDP options were used with QUIC, these limits would be applicable
> > to UDP options as well, i.e. 32 bytes of UDP options for IPv6, 52
> > bytes for IPv4 at least for initial packets and path validation.
>
> I think they missed something significant, even if just for IPv6 at 1500B.
>
> Joe
>
>