On Fri, Jul 09, 2010 at 07:25:45PM +0200, Philipp Leusmann wrote:
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: dict_tcp_lookup: send: get
> be...@xxx.de
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: dict_tcp_lookup: recv: 200
> DEFER%20User%20over%20quota
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: dict_tcp_lookup: found: DEFER
> User over quota
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: check_table_result:
> tcp:localhost:1337 DEFER User over quota be...@xxx.de
So far, so good.
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: >>> START Recipient address
> RESTRICTIONS <<<
It should never get here, the code in check_table_result() handles strings
starting with "DEFER <SPACE or TAB> ..." directly, without delegating
to the "generic_checks" code.
Either your Postfix source is modified, miscompiled, the binaries are
corrupted, or CPU is mal-functioning.
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: generic_checks: name=DEFER
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: NOQUEUE: reject: RCPT from
> mta-2.ms.rz.RWTH-Aachen.DE[134.130.7.73]: 450 4.3.2 <be...@xxx.de>: Recipient
> address rejected: Try again later; from=<philipp.leusm...@rwth-aachen.de>
> to=<be...@xxx.de> proto=ESMTP helo=<mta-2.ms.rz.rwth-aachen.de>
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: warning: restriction `User'
> after `defer' is ignored
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: generic_checks: name=DEFER
> status=2
> Jul 9 19:15:25 s15277780 postfix/smtpd[22232]: generic_checks:
> name=check_recipient_access status=2
>
> postconf -n shows this:
>
> smtpd_recipient_restrictions =
> permit_mynetworks,
> reject_unauth_destination,
> reject_unlisted_recipient,
> check_recipient_access tcp:localhost:1337
OK, this is an access(5) check as expected, and goes through
check_table_result(), which implements "DEFER <text>". So your system
is not behaving as the original Postfix would on a working machine
running correctly compiled code.
> BTW: I am running Postfix 2.5.5-1.1 on Debian Etch (?. I never can
> remember. The latest one :) )
The relevant code has not changed in quite some time.
--
Viktor.
