Date:  Sat, 2 Sep 2006 21:05:39 +0200
From:  Michael Stauber <bq (at mark) solarspeed.net>
Subject:  [coba-e:06695] Re: Dumbfounded
To:  coba-e (at mark) bluequartz.org
Message-Id:  <200609022105.40051.bq (at mark) solarspeed.net>
In-Reply-To:  <fc.000f8147002487fa000f8147002484fb.2487fb (at mark) fc.zio.com>
References:  <fc.000f8147002484fb000f8147002484fb.248500 (at mark) fc.zio.com> <200609020207.32091.bq (at mark) solarspeed.net> <fc.000f8147002487fa000f8147002484fb.2487fb (at mark) fc.zio.com>
X-Mail-Count: 06695

Hi Steve,

> I believe it is sendmail, the maillog tells me that sendmail rejects,
> almost immediately all aliases I test.
>
> Sep  2 09:44:28 ibm1 sendmail[29099]: NOQUEUE: connect from
> py-out-1112.google.com [64.233.166.177]
> Sep  2 09:44:28 ibm1 sendmail[29099]: AUTH: available mech=PLAIN ANONYMOUS
> LOGIN CRAM-MD5 DIGEST-MD5, allowed mech=EXTERNAL GSSAPI DIGEST-MD5 CRAM-MD5
> LOGIN PLAIN Sep  2 09:44:28 ibm1 sendmail[29099]: k82EiSH2029099: Milter:
> no active filter Sep  2 09:44:28 ibm1 sendmail[29099]: k82EiSH2029099:
> <stephen (at mark) zio.com>... User unknown
>
> I did create another server with no packages (MailScanner, ClamAV,
> SpamAssassin) and the alias work. I am of course about to re-add those
> packages and test the alias to see which one breaks sendmail.
>
> Any other suggestions before I total repunt, again?

On the "broken" box, do /etc/mail/virtusertable and/or /etc/mail/aliases look 
ok? 

Can you manually rehash the DB files?

makemap hash /etc/mail/aliases.db < /etc/mail/aliases
makemap hash /etc/mail/virtusertable.db < /etc/mail/virtusertable

I don't know which third party PKGs you have, but it is imagineable that they 
eventually may break Berkeley DB support, which then would make it impossible 
for Sendmail to read the *.db files. This is somewhat speculative, but worth 
investigating.

-- 

With best regards,

Michael Stauber