Message Doubles

[Drew Marshall]

On Wed, March 29, 2006 09:49, spart cus wrote:
> Here's one of the logs that produces the same email.
>
>  028e01c6537e$4aed2cc0$1a01a8c0 at globalknfb8aba
>  028301c6537e$3c71bf30$1a01a8c0 at globalknfb8aba
>
>  [root at mail log]# cat maillog |grep  globalknowledge at pldtdsl.net
>  Mar 29 13:02:01 mail postfix/cleanup[32433]: 236D8362639: hold: header
> Received\
>  : from exsmtp02.epldt.local (exsmtp02.epldt.biz [203.131.76.231])??by
> mail.drea\
>  m.ph (Postfix) with ESMTP id 236D8362639??for <care at dream.ph>; Wed, 29 \
>  Mar 2006 13:02:01 +0800 (PHT) from exsmtp02.epldt.biz[203.131.76.231];
> from=<gl\
>  obalknowledge at pldtdsl.net> to=<care at dream.ph> proto=ESMTP
> helo=<exsmtp02.ep\
>  ldt.local>
>  Mar 29 13:02:01 mail postfix/cleanup[32433]: 236D8362639: hold: header
> Received\
>  : from globalknfb8aba ([58.69.89.211]) by exsmtp02.epldt.local with
> Microsoft S\
>  MTPSVC(6.0.3790.211);?? Wed, 29 Mar 2006 12:50:59 +0800 from
> exsmtp02.epldt.biz\
>  [203.131.76.231]; from=<globalknowledge at pldtdsl.net>
> to=<care at dream.com.ph> pro\
>  to=ESMTP helo=<exsmtp02.epldt.local>
>  Mar 29 13:02:13 mail postfix/qmgr[2148]: 16C0F362643:
> from=<globalknowledge at pld\
>  tdsl.net>, size=31958, nrcpt=1 (queue active)
>  Mar 29 13:22:24 mail postfix/cleanup[327]: 3047D362639: hold: header
> Received: \
>  from exsmtp01.epldt.local (exsmtp01.epldt.biz [203.131.76.230])??by
> mail.dream.\
>  com.ph (Postfix) with ESMTP id 3047D362639??for <care at dream.ph>; Wed, 29
> Ma\
>  r 2006 13:22:24 +0800 (PHT) from exsmtp01.epldt.biz[203.131.76.230];
> from=<glob\
>  alknowledge at pldtdsl.net> to=<care at dream.ph> proto=ESMTP
> helo=<exsmtp01.epld\
>  ----:%%-F1  Untitled 1~
> (Fundamental)--L1--Top----------------------------

These look like seperate messages. Look at the times in the logs, 20
minutes a part. A better check would be if I could see the whole log for
13:02:01 -> 13:04:00 and 13:22:24 -> 13:24:00

This should show the first delivery and then the second. The problem is
that Postfix can and does reuse it's message ids as they are based on
inode numbers. If your /var/spool/postfix partition is a bit small/ full
it will reuse the numbers faster. The theory goes that Cyrus should be
using unique message ids so there is no duplication. I know Postfix
generates another id when it delivers the mail it's self (To unix mailbox
or maildir) as opposed to handing off the the Cyrus delivery agent.

Drew


-- 
In line with our policy, this message has 
been scanned for viruses and dangerous 
content by MailScanner, and is believed to be clean.
www.themarshalls.co.uk/policy