[374] in netbsd-help mailing list archive
Re: daily report...
daemon@ATHENA.MIT.EDU (John Hawkinson)
Sun Oct 22 10:33:18 1995
Date: Sun, 22 Oct 1995 10:33:02 -0400
To: I-Lin Wang <ilin@MIT.EDU>
Cc: netbsd-help@MIT.EDU
In-Reply-To: "[364] in netbsd-help mailing list archive"
From: John Hawkinson <jhawk@MIT.EDU>
> Date: Wed, 18 Oct 1995 13:17:39 EDT
Sorry about the delay; inexcusable as always.
> I found there are some strange messages in my daily output
> email like:
> mail:
> Mail Queue (4 requests)
> --Q-ID-- --Size-- -----Q-Time----- ------------Sender/Recipient------------
> LAA23704 286 Fri Oct 13 11:20 ilin
> (Deferred: Connection refused by japan.mit.edu.)
> pmberger@japan.mit.edu
> OAA02600 1096 Fri Sep 22 14:51 ilin
> (Deferred: Connection refused by emin05.mail.aol.com.)
> FelipeChen@aol.com
> XAA03010 1506 Wed Sep 20 23:09 ilin
> (reply: read error from cc.nsysu.edu.tw.)
> ayusg@cc.nsysu.edu.tw (An-Yeong Sun)
> LAA23695 2167 Fri Oct 13 11:11 ilin
> (Deferred: Connection refused by japan.mit.edu.)
> pmberger@japan.mit.edu
> Does that mean those emails weren't sent properly?(i.e. the
> receiver did not receive these mails??)
Indeed. There is apparently a bug in the installation procedure
such that the sendmail queue running process is not properly
added to crontab.
To explain what this means:
If for some reason mail delivery to a particular site failes
(the site is down, too busy, out of disk space, transient
network failure, whatever), mail is queued on the sending machine
for delivery at a later time. Periodically, the queue is "run",
meaning that for each message in the queue, sendmail attempts
to redeliver the message and remove it from the queue.
To manually run the queue, you should (as root) run:
/usr/sbin/sendmail -v -q
(-v will cause it to show you the incremental progress).
To fix this problem, as root, please run crontab -e, and add
the following line to your crontab after the atrun line:
*/30 * * * * /usr/sbin/sendmail -q
This will cause sendmail to run the queue every 30 minutes.
We apologize for this screw up, and I'll try and figure out where
the install procedure went astray and fix it up. Thanks for
noting this.
--jhawk@mit.edu
John Hawkinson
