Am 02.10.2014 um 16:27 schrieb Daniele Nicolodi: > On 02/10/14 16:22, [email protected] wrote: >> Am 02.10.2014 um 16:16 schrieb [email protected]: >>> Am 02.10.2014 um 16:10 schrieb Daniele Nicolodi: >>>> I have a system with a few local users where some of them are configured >>>> for forwarding all incoming messages to external addresses via the use >>>> of .forward. >>>> >>>> One of those users forwards mail to an hotmail.com address. >>>> >>>> When mail is delivered to this address this is what happens: >>>> >>>>> Oct 2 13:50:58 zed postfix/smtpd[1063]: connect from >>>>> bay004-omc1s27.hotmail.com[65.54.190.38] >>>>> Oct 2 13:50:59 zed postfix/smtpd[1063]: NOQUEUE: reject: RCPT from >>>>> bay004-omc1s27.hotmail.com[65.54.190.38]: 450 4.1.8 >>>>> <[email protected]>: Sender address >>>>> rejected: Domain not found; >>>>> from=<[email protected]> >>>>> to=<[email protected]> proto=ESMTP helo=<BAY004-OMC1S27.hotmail.com> >>>>> Oct 2 13:50:59 zed postfix/smtpd[1063]: disconnect from >>>>> bay004-omc1s27.hotmail.com[65.54.190.38] >> >> BTW what you show here is *not* "One of those users forwards mail >> to an hotmail.com address" because the connection comes *from* >> hotmail and is pretty sure rejected by "reject_unknown_sender_domain" > > Please stop chopping up my email to make it read differently, the first > lines of the log are clearly an email received for [email protected] and > forwarded to [email protected]. > The connection to the delivery attempt from accounts.microsoft.com to > grinta.net is correlated to the forwarding only by its timing
what makes you sure that your outgoing mail and the incoming connection are related? for that evidence it needs to happen almost every time that it is repeated at least is caused by your 450 response maybe the sender domain has enforced DMARC and must not be forwarded > Should I do something about it? if you are interested and that is repeatable just disable the restriction temporary and look into that mail, otherwise leave it enabled, the invalid sender is not your fault - could be sender-verification with a wrong configured probe-sender on the other side and since you answer with 450 it is repeated
