Maxime Gaudreault:
> Good answer ;)
>
> But, I just did some test, and more than 24 hours after the I first
> noticed the issue, it still relay emails through the default relayhost:
No, it doesn't.
> Jun 12 16:43:03 mx01 postfix/smtpd[30843]: connect from
> mailrelay7.megamailservers.com[209.235.143.13]
> Jun 12 16:43:13 mx01 postfix/smtpd[30843]: NOQUEUE: reject: RCPT from
> mailrelay7.megamailservers.com[209.235.143.13]: 550 5.1.1
> <recipi...@domain.ca>: Recipient address rejected: undeliverable
> address: host 10.0.1.11[10.0.1.11] said: 550 not local host domain.ca,
> not a gateway (in reply to RCPT TO command);
This is an OLD result from your address verification cache.
Apparently, you have a corrupted cache file, or you have messed
with the address_verify expiration times.
Remove the cache file and do "postfix reload".
> Jun 12 16:43:13 mx01 postfix/cleanup[30845]: B5281560068:
> message-id=<20120612204313.b5281560...@mx01.reference.qc.ca>
> Jun 12 16:43:13 mx01 postfix/qmgr[7521]: B5281560068:
> from=<double-bou...@reference.qc.ca>, size=246, nrcpt=1 (queue active)
> Jun 12 16:43:13 mx01 postfix/smtp[30902]: B5281560068: to=<
> recipi...@domain.ca >, relay=192.168.100.9[192.168.100.9]:25,
> delay=0.01, delays=0/0/0/0, dsn=2.1.5, status=deliverable (250 2.1.5
> Recipient < recipi...@domain.ca > ok (local))
This is a NEW ADDRESS VERIFICATION PROBE that is sent to the right
relay host.
Wietse
>
> And it's not all emails that goes through the wrong server..
>
> -----Original Message-----
> From: wie...@porcupine.org [mailto:wie...@porcupine.org]
> Sent: June-12-12 2:34 PM
> To: Maxime Gaudreault
> Cc: wie...@porcupine.org
> Subject: Re: Postfix sometime ignoring transport map
>
> Maxime Gaudreault:
> > How do you know it came from the cache ?
>
> because I wrote the code that produced that message in your logfile.
>
> > How do I reset this cache ?
>
> This is not needed. The bad entry expires after four hours.
>
> Wietse
> > Max
> >
> >
> > -----Original Message-----
> > From: wie...@porcupine.org [mailto:wie...@porcupine.org]
> > Sent: June-12-12 11:55 AM
> > To: Maxime Gaudreault
> > Cc: Postfix users
> > Subject: Re: Postfix sometime ignoring transport map
> >
> > Maxime Gaudreault:
> > > Jun 11 17:07:00 mx01 postfix/smtp[2003]: 61C50561039: to=<
> > > recipi...@domain.ca>, relay=192.168.100.9[192.168.100.9]:25,
> > > delay=3.4, delays=3.4/0/0/0, dsn=2.0.0, status=sent (250 2.0.0
> > > 4fd65e47-000321f5 Message accepted for delivery)
> >
> > Wietse:
> > > This appears to be a real email message.
> >
> > Maxime Gaudreault:
> > > from=<double-bou...@reference.qc.ca>, size=246, nrcpt=1 (queue
> > > active)
> >
> > > Jun 11 16:56:15 mx01 postfix/smtp[800]: 2093E56006D:
> > > to=<recipi...@domain.ca>, relay=192.168.100.9[192.168.100.9]:25,
> > > delay=0.01, delays=0/0/0/0, dsn=2.1.5, status=deliverable (250 2.1.5
>
> > > Recipient <recipi...@domain.ca> ok (local))
> >
> > Wietse:
> > > This appears to be an "address verification" message.
> > >
> > > Both messages are routed via relay=192.168.100.9[192.168.100.9]:25.
> > > Why is one delivery "good", and one delivery "bad"?
> >
> > Maxime Gaudreault:
> > > Bad delivery tries to send through 10.0.1.11 but verify the
> > > recipient with 192.168.100.9.
> > > 10.0.1.11 can't accept mail for this domain
> >
> > The logfile evidence shows that BOTH ARE ROUTED VIA 192.168.100.9.
> >
> > The REJECT message involving 10.0.1.11 comes from the address
> > verification CACHE. This CACHE uses historical information.
> > Apparently, you made some configuration error that has been fixed in
> > the meantime.
> >
> > Wietse
> >
> >
>
>
>
