On Mon, 23 Dec 2019 at 22:01, Ray Bellis <[email protected]> wrote:
>
> On 20/12/2019 15:08, Bob Harold wrote:
>
> > But if we are updating it, could we consider a better word than
> > "forward" ? Actually "backward" would be correct, although I prefer
> > "from the back to the front" as used elsewhere.
>
> It's not possible to traverse the RRs in a raw DNS packet "backwards".
> You have to start at the beginning and remember the offset of each RR
> found.
>
Is there an implicit assumption here that all RRs beyond the UDP limit
are to be discarded without requiring inspection of their internal
content?
If so, then I have a problem with the current EDE draft which appears
to involve preferentially discarding the EDE option from the OPT RDATA
(as distinct from the entire OPT RR).
draft-ietf-dnsop-extended-error-13 says:
3. Extended DNS Error Processing
When the response grows beyond the requestor's UDP payload size
[RFC6891], servers SHOULD truncate messages by dropping EDE options
before dropping other data from packets. Implementations SHOULD set
the truncation bit when dropping EDE options.
--
RWF
_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop
