On Thu, 6 May 1999, Johann Spies wrote:
> I have a dialup system to my ISP ans is using exim as mailagent on a
> Debian 2.0 system.
Which release of Exim, please?
> What is strange to me is that this error is very "selective" eg. I sent
> mail to a list of 41 people and when I do a mailq a few minutes later
> (before I dialed in to the ISP) the following shows (the list is shortened
> to save some space):
>
> 1.7K 10f0Yy-0001iX-00 <jhspies@???>
> aabasson@???
> D CMarais@???
> D HJBC@???
> conni@???
> D johanzac@???
> D vdberg@???
> barry1@???
>
> Only the messages with the D got the time-out error and I have no idea
> why.
That can't be right. The 'D' means "delivered". If you haven't dialed to
the ISP, it can't have delivered the message. Oh, wait, it could have it
it has bounced it. Hmm, yes...
> 1999-05-05 14:14:10 10f0Yy-0001iX-00 == CMarais@??? R=smarthost
> defer (-1):
> 1999-05-05 14:14:10 10f0Yy-0001iX-00 ** CMarais@???: retry timeout
> exceeded
It seems as though you may have got a retry time set for your smarthost
which has expired. If you are running a default configuration file, it
will try to deliver the message immediately, and fail (because you
haven't dialled up). That will cause a retry time to be set. Exim has to
be tweaked in various ways if you use it on intermittently connected
hosts (it wasn't designed for that environment, but people seem to be
able to use it). Try
(1) Deleting all the files in the /var/spool/exim/db directory - that
will wipe out all the retry data and start from a clean slate.
(2) Set "queue_remote" in your Exim configuration. That will stop it
trying to send out mail to remote hosts until you tell it to.
(3) When you have dialled in to your ISP, run "exim -qqf" to get it to
route the messages, and then deliver them.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
