T_ANY

[Mark Andrews]

In message <alpine.LSU.2.00.1003202112330.2053 at hermes-2.csi.cam.ac.uk>, Tony Fi
nch writes:
> On Sat, 20 Mar 2010, Glenn English wrote:
> >
> > Just why qmail reports a T_ANY failure as a CNAME failure, I also don't
> > know.
> 
> This is a bug in qmail. It tries to canonicalize domains in the SMTP
> envelope of outgoing messages. It originally did this by performing CNAME
> queries on each domain, but this caused interop problems with BIND 4.

No it didn't.  BIND 4 always answered CNAME queries correctly if
the zone content was correct.  What happened was people didn't correct
load errors and a ANY query worked around the fact that they didn't correct
the loadtime errors.

> The
> CNAME query was changed to an ANY query, which is buggy in a number of
> ways: ANY queries are much more likely to trigger qmail's undersized DNS
> buffer bug; and ANY queries are supposed to be used for debugging not for
> normal operations.
> 
> The correct way to canonicalize a mail domain is to do an MX query (or
> failing that A and AAAA queries) and if the domain is not canonical the
> response will include the CNAME record.
> 
> The reason for canonicalization was conformance with RFC 1123 section
> 5.2.2. However when qmail was written this was not necessary for
> interoperability and the requirement was removed in the DRUMS process
> that yielded RFC 2821.
> 
> So the way to actually fix the bug is to delete the canonicalization code.
> 
> Tony.
> -- 
> f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/
> GERMAN BIGHT HUMBER: SOUTHWEST 5 TO 7. MODERATE OR ROUGH. SQUALLY SHOWERS.
> MODERATE OR GOOD.
> _______________________________________________
> bind-users mailing list
> bind-users at lists.isc.org
> https://lists.isc.org/mailman/listinfo/bind-users
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org