On Wed, Apr 20, 2016, at 11:08 AM, Wietse Venema wrote:
> SENDMAIL(1) SENDMAIL(1)
>
> ...
> -bv Do not collect or deliver a message. Instead, send an email
> report after verifying each recipient address. This is useful
> for testing address rewriting and routing configurations.
Thanks.
Probes with 'sendmail -bv' work, from my shell, with no error. :-/
Since it's not identical to the path/process an inbound mail takes, which I
think is what this means
> Actually, it's a different fake delivery status notification message
> than with address probes; it becomes a real email message. Verify
> probe results are absorbed internally.
I susepct it's not touching whatever I've done to cause my problem.
I turned up debug logging (wow, noisy!), and am staring at this bit which IIUC
is the address verify in progress. Notice the ERROR,
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: connect to subsystem
private/verify
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: send attr request =
query
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: send attr address =
exampl...@example.com
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: private/verify
socket: wanted attribute: status
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
name: status
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
value: 0
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: private/verify
socket: wanted attribute: recipient_status
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
name: recipient_status
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
value: 2
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: private/verify
socket: wanted attribute: reason
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
name: reason
>> Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
>> value: mail for example.com loops back to myself
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: private/verify
socket: wanted attribute: (list terminator)
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: input attribute
name: (end)
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: NOQUEUE:
reject_warning: RCPT from out1-smtp.messagingengine.com[66.111.4.25]: 550 5.1.1
<exampl...@example.com>: Recipient address rejected: undeliverable address:
Recipient address verification failed; from=<ja...@2.example.com>
to=<exampl...@example.com> proto=ESMTP helo=<out1-smtp.messagingengine.com>
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: generic_checks:
name=reject_unverified_recipient status=0
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: generic_checks:
name=reject_unlisted_recipient
Apr 20 11:02:17 mail01 postfix/psint/smtpd[24202]: >>> CHECKING
RECIPIENT MAPS <<<
That
input attribute value: mail for example.com loops back to myself
can't be good. I wonder if that's what's causing the almost immediately
subsequent "Recipient address rejected" (for now, now relegated to just a
reject_warning)?
What I can't yet grasp is which piece is causing the problem.
If what Noel said is true
The "loops back to myself" message occurs when postfix is greeted with
its own HELO hostname.
there's a wrong HELO, or hostname setting ... somewhere. Since it's an
'internal' process, I can't seems to identify where the problem is. The
increased verbosity on the psint process isn't enough to tell me, apparently.
Where do I look for this cause?
Jason
