Corrupted messages Postfix

[Glenn Steen]

Hello David,

This problem you have may be stemming from the change of locking
scheme in Postfix 2.10. It was also a problem (in the past) when we
did the two-postfix deferred queue thing to implement MailScanner
(prior to the hold queue thing we usually do now).
Check wether your postfix is using fifo or unix locks in master.cf,
and consider going back to fifo in that (with appropriat restart of
Postfix afterwards, of course)... OR change MailScanner.conf
accordingly and restart MailScanner (setting is "Lock Type").

I would think it fairly safe&easy to do the PF change.

Cheers
-- 
-- Glenn

On 31 May 2013 09:18, David Valin Alonso <deivishome at gmail.com> wrote:
> Hi Martin,
> this is a copy/pste from my mail.log:
> May 29 11:02:21 server postfix/cleanup[27995]: 19D6CDFD3F:
> message-id=<201305290411.3.5607.30342.2445274 at cog.lumata.com>
> May 29 11:02:21 server postfix/smtpd[27991]: disconnect from
> unknown[213.92.42.10]
> May 29 11:02:22 server MailScanner[24390]: New Batch: Scanning 1 messages,
> 4532 bytes
> May 29 11:02:22 server MailScanner[24390]: Virus and Content Scanning:
> Starting
> May 29 11:02:36 server MailScanner[24390]: Requeue: 19D6CDFD3F.AEAD7 to
> C6CEEDFD46
> May 29 11:02:36 server MailScanner[24390]: Uninfected: Delivered 1 messages
> May 29 11:02:36 server postfix/qmgr[3507]: C6CEEDFD46: from=<>, size=3885,
> nrcpt=1 (queue active)
> May 29 11:02:36 server postfix/qmgr[3507]: warning: C6CEEDFD46: message
> rejected: missing end record
> May 29 11:02:36 server postfix/qmgr[3507]: warning: saving corrupt file
> "C6CEEDFD46" from queue "active" to queue "corrupt"
> May 29 11:02:36 server MailScanner[24390]: Deleted 1 messages from
> processing-database
>
> Not all mails go to corrupt, for example if a mail comes to 1 person i
> handles well the first time, the next time it sends to corrupt, i am loosing
> the 60-75% in the corrupt queue, yesterday i had to stop MailScanner and
> reconfig headers_checks to bypass the problem till i find a solution because
> everything was working great.
>
> Regards,
>
> David
> 2013/5/30 Martin Hepworth <maxsec at gmail.com>
>>
>> what do the logs say for the messages in question?
>> check the postfix logs and the mailscanner logs.
>> Also try running mailscanner in debug mode (see the wiki)
>>
>> --
>> Martin Hepworth, CISSP
>> Oxford, UK
>>
>>
>> On 30 May 2013 09:10, David Valin Alonso <deivishome at gmail.com> wrote:
>>>
>>> Hello,
>>> i got a server runing ubuntu 10.04 lts x64 + postfix 2.10 + cyrus 2.2 +
>>> Mailscanner 4.84 + Spamassasin and all was working really great til a couple
>>> of days that began to send mails to corrupt folder
>>> /var/spool/postfix/corrupt. It complains about a missing record and it
>>> rejects the mail moving from active to corrupt queue.
>>> I don't know what happend and the configurations of postfix and
>>> mailscanner didn't change as it was working really great.
>>>
>>> What could happened?
>>>
>>> Regards,
>>>
>>> David
>>>
>>> --
>>> MailScanner mailing list
>>> mailscanner at lists.mailscanner.info
>>> http://lists.mailscanner.info/mailman/listinfo/mailscanner
>>>
>>> Before posting, read http://wiki.mailscanner.info/posting
>>>
>>> Support MailScanner development - buy the book off the website!
>>>
>>
>>
>> --
>> MailScanner mailing list
>> mailscanner at lists.mailscanner.info
>> http://lists.mailscanner.info/mailman/listinfo/mailscanner
>>
>> Before posting, read http://wiki.mailscanner.info/posting
>>
>> Support MailScanner development - buy the book off the website!
>>
>
>
> --
> MailScanner mailing list
> mailscanner at lists.mailscanner.info
> http://lists.mailscanner.info/mailman/listinfo/mailscanner
>
> Before posting, read http://wiki.mailscanner.info/posting
>
> Support MailScanner development - buy the book off the website!
>



-- 
-- Glenn
email: glenn < dot > steen < at > gmail < dot > com
work: glenn < dot > steen < at > ap1 < dot > se