Index: [Article Count Order] [Thread]

Date:  Wed, 04 Jun 2008 11:48:28 -0400
From:  Richard Siddall <cobalt (at mark) elirion.net>
Subject:  [coba-e:13156] Re: Another Strange Sendmail Error
To:  coba-e (at mark) bluequartz.org
Message-Id:  <4846B94C.20104 (at mark) elirion.net>
In-Reply-To:  <772066.26924.qm (at mark) web65611.mail.ac4.yahoo.com>
References:  <772066.26924.qm (at mark) web65611.mail.ac4.yahoo.com>
X-Mail-Count: 13156

Dan Kriwitsky wrote:
> I did not enter the contents via the GUI. I copy the access file from
> /etc/mail/access, download it as ASCII, edit it, upload it as ASCII, copy
> it back to /etc/mail/ and run makemap hash access.db < access.
> 

I was suggesting editing it via the GUI just to confirm that it's the 
file sendmail is using.  (I'm assuming the BQ GUI will edit the right file.)

> -rw-r--r--   1 root  root   915043 Jun  4 10:07 access
> -rw-r-----   1 smmsp root  2531328 Jun  4 10:07 access.db
> 

That pretty much confirms it's the right file, since sendmail is 
updating its access.db from it.

> When running tail -f /var/log/maillog I will see go by:
> Jun  4 08:46:35 admin sendmail[18758]: ruleset=check_relay,
> arg1=236-145.chk.adsl.eastport.ru, arg2=82.179.236.145,
> relay=236-145.chk.adsl.eastport.ru [82.179.236.145], reject=550 5.0.0 mail
> from RU no longer accepted
> 
> Because I have in access:
> ru[TAB]550 mail from RU no longer accepted
> 
> But, I also have:
> ua[TAB]REJECT
> And just now, spam came through from there:
> Return-Path: <ybuw (at mark) bouvetcousin.com>
> Received: from 12-60dabc5bb0af (89.252.2.220.freenet.com.ua [89.252.2.220]
> (may be forged))
> 	by sub.example.com (8.13.1/8.13.1) with ESMTP id m54DVAo1018485
> 	for <me (at mark) example.com>; Wed, 4 Jun 2008 08:31:12 -0500
> 
> That should have been rejected by SMTP upon connect from *.ua.
> 

Is there anything in the access file between those two lines that might 
cause makemap to give up on the file?

Is there anything at the end of the "ua" line - unprintable characters?
(I'm grasping at straws here.)

> It's certainly frustrating. This always worked just fine on the RaQ and
> editing any other text configuration file offline this way works just fine.
> 

If it was easy to dump the .db file and I was sitting at the machine, 
I'd probably browse through the dump to see if that revealed any clues.

Regards,

	Richard Siddall