Hi Andrew,

Thank you for your very helpful comments.  In particular, you said:

> when you go back on-line, then the dns lookups can work again, so
> the queue can be flushed.  

Sometimes I go on line just for a moment and just for the purpose of
sending mail.  So, the solution apparently is to "wake up" the 
queue runner and get it to flush the clientmqueue as soon as I go on 
line.  What would be the best way to to sound the reveille bugle?

Thanks again!

Kindest regards,

Tom

On Mar 31, 10:19am, Andrew Brown wrote:
} Subject: Re: new sendmail spooling and dns
} >I noticed that there was a change in the headers of the daily output
} >and the daily insecurity output messages in root's mailbox.  This 
} >change happened when I rebuilt -current and installed the new sendmail.  
} >Maybe this change will give somebody the information needed to clue 
} >me in on why I am getting DNS timeout warning messages from the 
} >mailer daemon respecting the messages that new sendmail can't seem to 
} >move out of /var/spool/clientmqueue.  
} 
} you're getting the dns timeout because you're off-line.  when you're
} off-line, sendmail can't look up (or verify) the existence of the
} address to which (or from which) you are sending the mail, so it stays
} in the client queue.  when you go back on-line, then the dns lookups
} can work again, so the queue can be flushed.
} 
} this is perfectly normal behavior.
} 
} >Here are the headers from before the upgrade:
} >
} >>From daemon Sat Mar 29 03:17:03 2003
} >Return-Path: <root>
} >Received: (from root@localhost)
} >	by aluminum.miller.net (8.11.6/8.11.6) id h2T8F0b05932;
} >	Sat, 29 Mar 2003 03:15:00 -0500 (EST)
} >Date: Sat, 29 Mar 2003 03:15:00 -0500 (EST)
} >From: Charlie Root <root>
} >Message-Id: <200303290815.h2T8F0b05932@aluminum.miller.net>
} >To: root
} >Subject: aluminum.miller.net daily output for Sat Mar 29 03:15:00 EST 2003
} >Status: OR
} 
} when you were using 8.11.6, the binary was suid root and used one
} queue for all mail.
} 
} >Here are the headers after the upgrade:
} >
} >>From tmiller@flex.com Sun Mar 30 03:17:04 2003
} >Return-Path: <tmiller@flex.com>
} >Received: from aluminum.miller.net (localhost [127.0.0.1])
} >	by aluminum.miller.net (8.12.8/8.12.8) with ESMTP id h2U8H41R002829
} >	for <root@aluminum.miller.net>; Sun, 30 Mar 2003 03:17:04 -0500 (EST)
} >Received: (from root@localhost)
} >	by aluminum.miller.net (8.12.8/8.12.8/Submit) id h2U8F0HU000663;
} >	Sun, 30 Mar 2003 03:15:00 -0500 (EST)
} >Date: Sun, 30 Mar 2003 03:15:00 -0500 (EST)
} >From: Charlie Root <tmiller@flex.com>
} >Message-Id: <200303300815.h2U8F0HU000663@aluminum.miller.net>
} >To: root@aluminum.miller.net
} >Subject: aluminum.miller.net daily output for Sun Mar 30 03:15:00 EST 2003
} >Status: OR
} >
} >Any clues?
} 
} now that you're using 8.12.8, the binary is sgid to the smmsp group,
} and uses two queues: one for local collection and one for everything
} else (into which the locally collected mail is moved by way of smtp
} between the process you ultimately started and the smtp daemon
} actually running as root).
} 
} so you have two "received:" headers because two sendmail processes
} handled the mail.  note that the first one (the bottom-most one in
} your sample above) says "8.12.8/8.12.8/Submit" and the other does not.
} 
} -- 
} |-----< "CODE WARRIOR" >-----|
} codewarrior@daemon.org             * "ah!  i see you have the internet
} twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
} werdna@squooshy.com       * "information is power -- share the wealth."
} 
}-- End of excerpt from Andrew Brown