James Brown wrote:
On 26/09/2008, at 1:01 AM, Wietse Venema wrote:
James Brown:
Examining the headers of the email I sent to this list:
1. Received: from [192.168.1.10] ([127.0.0.1]
helo=[192.168.1.10]) by
...
3. Received: from mail.bordo.com.au ([127.0.0.1]) by localhost
(mail.bordo.com.au [127.0.0.1]) (amavisd-new, port 10024) with ESMTP
id pW2vMuVVzAvF for <postfix-users@postfix.org>; Thu, 25 Sep 2008
11:30:44 +1000 (EST)
...
8. Received: from mail.bordo.com.au ([127.0.0.1]) by localhost
(mail.bordo.com.au [127.0.0.1]) (amavisd-new, port 10024) with ESMTP
id pW2vMuVVzAvF for <postfix-users@postfix.org>; Thu, 25 Sep 2008
11:30:44 +1000 (EST)
Are you sure that the exact same header is logged twice, with
four other Received: headers in-between?
Wietse
Don't you see the same thing in the headers to the list that I see?
James.
It's quite unlikely that amavisd-new would scan the same
message twice within the same second using the same ID number.
It's also quite unlikely that amavisd-new is adding multiple
headers in random places.
Postfix has never been known to rearrange or duplicate headers.
It's far more likely that something downstream of postfix and
amavisd-new is duplicating headers. and maybe it mangles some
of them for good measure...
Hmmm, what could be downstream of postfix & amavisd-new...
Sounds as if what you really need is a tcpdump of a message
entering your dkim signing gateway and of the same message
leaving.
At the least, you should HOLD or defer some message on the
postfix box to examine before it leaves postfix.
A transport entry like
[EMAIL PROTECTED] retry:
will defer all mail addressed to that recipient.
Then you can see if the duplicated/mangled headers exist
before the mail leaves postfix.
--
Noel Jones
