spf-discuss
[Top] [All Lists]

Re: DNS load research

2005-03-25 13:17:15
Radu Hociung wrote:

It's completely up to the DNS server implementation whether
it sends information it wasn't asked for

Okay.  OTOH Wayne _has_ the separate limit of 10 MXs per q=mx,
so I guess that all our DNS discussions were futile:  you can
count mechanisms and redirect=, as Wayne does it, this doesn't
depend on the interal behaviour of DNS, and you can exclude
very expensive cases of q=mx or ptr-stuff by a separate limit,
as Wayne does it.

The limits are fine as they are.  But we could still move to an
"abstract query limit", where q=mx is counted as 1 + MXs, dito
PTR.  But that's not necessarily related to the real number of
DNS queries, it's only an abstraction to get the same view of
a sender policy in "setup wizards", "validators", etc.

there may not be enough room in one UDP packet the IP
addreses of those hosts, and maybe not even enough room for
all the names.

For the latter case we would need another "MAY ignore nonsense
before trying TCP", same idea as for a complete SPF or TXT RR.

it would seem that that A mechanism is also not as reliable
as we believe it to be, in the SPF application.

That's a known problem, I've discussed it some months ago with
a case like news.clara.net on spf-help:

195.245.201.150  sabbath.news.uk.clara.net
195.245.201.151  spandrell.news.uk.clara.net
217.158.240.10   eunomia.uk.clara.net
217.158.240.11   echo.uk.clara.net
217.158.240.12   lotis.uk.clara.net
217.158.240.20   nnrp-t71-01.news.uk.clara.net
217.158.240.23   nnrp-t71-02.news.uk.clara.net
217.158.240.24   nnrp-t71-03.news.uk.clara.net
195.8.68.205     doris.uk.clara.net
195.8.68.206     dyke.uk.clara.net
195.8.68.207     iris.uk.clara.net
195.8.68.209     despina.uk.clara.net
195.8.68.217     ersa.uk.clara.net
195.8.68.218     demeter.uk.clara.net
195.8.68.222     damia.uk.clara.net

And I found the header file with a hardwired limit for the IPs,
but I forgot what it was.

Maybe these are some of the reasons why the DNS guys didn't
like SPF ?

They didn't like q=ns for the "zone cut", and of course they
don't like TXT for this purpose.  Both problems are solved (as
far as possible), no more "zone cut" in draft -01.

                           Bye, Frank



<Prev in Thread] Current Thread [Next in Thread>