[Mailman-Users] duplicates

Mark Sapiro mark at msapiro.net
Wed Oct 28 00:49:38 EDT 2015


On 10/27/2015 12:30 AM, Marco Stoecker wrote:
> On 10/26/2015 11:32 PM, Mark Sapiro wrote:
>> On 10/26/2015 02:14 PM, Marco Stoecker wrote:
>>
>>> I did attach the header from a recent message I got last week. I'm a
>>> member of a
>>> list and the sender sent this mail to 5 mailinglists on our server. Each
>>> member of these 5 lists got 5 messages.
>>>
>>> But the email, which I sent last week, with the attachement never made
>>> it to this list. How can I send the headers instead, cause this email
>>> would be very long?
>>
>>
>> I never saw such a message, even in the moderation queue because of size.


I and the list have now received it. It was sent during the recent
server outage on only reached the server earlier today.

It is helpful. I will copy some of the header info below.


...
>>
>> The most helpful information would be the Postfix log entries from
>> several minutes before Oct 22 10:40:33 2015 up to Oct 22 10:44:28 2015,
>>
> This is the related mail.log entry:
> 
> Oct 22 10:40:05 wakis02 postfix/smtpd[4898]: connect from localhost[::1]
> Oct 22 10:40:05 wakis02 postfix/smtpd[4898]: BF9AD1C94:
...


This doesn't start early enough. The duplication occurs at 10:37:47. The
duplication occurs because of the way your mail is ultimately delivered
to mailman. Here is an excerpt from header1.txt in your other mail:

> Received: from imap.1und1.de [212.227.15.188]

>  by wakis02.local with IMAP (fetchmail-6.3.26)

>  Thu, 22 Oct 2015 10:40:43 +0200 (CEST)

> Received: from [212.227.15.41] ([212.227.15.41]) by mx.kundenserver.de

>  (mxeue002) with ESMTPS (Nemesis) id 0M1foo-1aeGtV2fLW-00tjIT for

>  <mailmanserver at waldorfkindergarten-erlangen.de>; Thu, 22 Oct 2015 10:37:47

>  +0200

> Received: from mout.web.de ([212.227.15.3]) by mx.kundenserver.de (mxeue002)

>  with ESMTPS (Nemesis) id 0MVE1V-1a1cd42azt-00YP1m for

>  <ak-leiter at waldorfkindergarten-erlangen.de>; Thu, 22 Oct 2015 10:37:47 +0200

> Received: from Klamotten ([84.168.195.183]) by smtp.web.de (mrweb003) with

>  ESMTPSA (Nemesis) id 0MTh7A-1Zy14E1g36-00QRsw; Thu, 22 Oct 2015 10:37:47

>  +0200

> From: "Alexandra Kick" <alexandrakick at web.de>

> To: "'Alexandra Kick'" <alexandrakick at web.de>,

>  <vorstand at waldorfkindergarten-erlangen.de>,

>  <kollegium at waldorfkindergarten-erlangen.de>,

>  <ak-leiter at waldorfkindergarten-erlangen.de>,

>  <gruppensprecher at waldorfkindergarten-erlangen.de>,

>  <verwaltung at waldorfkindergarten-erlangen.de>,

>  <beisitz at waldorfkindergarten-erlangen.de>


Looking at the Received headers in chronological order (bottom to top),
the message is received by smtp.web.de and relayed (as mout.web.de) to
mx.kundenserver.de for the <ak-leiter at waldorfkindergarten-erlangen.de>
list. It is then relayed as is probably all list mail on that server to
itself for <mailmanserver at waldorfkindergarten-erlangen.de>. It is then
picked up by fetchmail and processed further.

The other messages headers all look the same except they are each
initially for one of the other lists, but they all get forwarded to
<mailmanserver at waldorfkindergarten-erlangen.de>.

Now, I'm sure what happens is one message for
<ak-leiter at waldorfkindergarten-erlangen.de> gets to
<mailmanserver at waldorfkindergarten-erlangen.de> and at some point later
the fact that it was originally received just for the
<ak-leiter at waldorfkindergarten-erlangen.de> list is lost or ignored and
the process looks at the To: header of the mail, sees 5 lists and
forwards the mail to all 5 lists.

The same thing happens with each of the other 4 messages resulting in 5
messages to each of the 5 lists.

The answer is your process for delivering mail to Mailman is flawed
because it takes a message received for one list only and delivers it to
every list mentioned in To: (and maybe Cc:).

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan


More information about the Mailman-Users mailing list