Re: [EXIM] [patrik@ein.cz: Bug#39726: exim: The retry rules …

Top Page
Delete this message
Reply to this message
Author: Dr Andrew C Aitchison
Date:  
To: exim-users
CC: Mark Baker, patrik
Subject: Re: [EXIM] [patrik@ein.cz: Bug#39726: exim: The retry rules vs. dial up connection.]

On Fri, 18 Jun 1999, Mark Baker wrote:
> ----- Forwarded message from Patrik Rak <patrik@???> -----
> The retry logic of exim (at least the one caused by eximconfig retry rule)
> is not very well designed for people who connect to the Internet only occasionally.
>
> For example, assume I write an email to someone and I don't mind if it stays
> in the queue until I use my dial up next time. If I forget to call for some
> time, I get the warning messages and then finally the mail is returned.
> OK, that's not the problem. The real problem is caused by the following exim
> behavior (quoted from oview.txt):
>
>     When the maximum time for retrying has passed,
>     pending addresses are failed.  However, a next try
>     time is still computed from the final subrule.
>     Until that time is reached, any new messages for
>     the address are immediately failed.  When the next
>     try time is passed, one further delivery attempt is
>     made; if this fails, a new next try time is
>     computed, and so on.

>
> You see? This means that once I get the mail bounced, all mail for that
> address is bounced immediately.


On Fri, 18 Jun 1999, Jim Knoble replied:
> Ummm ... this isn't particularly the author's fault, is it? It's easy
> enough to configure the retry period to be longer. For instance the
> following retries every ten minuets for 30 days:
>
>     *         *               F,30d,10m

>
> This appears to be a configuration issue. Have i missed something?


I have some sympathy with the original complaint.
Under many circumstances, sending a 3 week old mail may not be a good
idea, but you do want a new message to go straight out.

As Phil has mentioned, he doesn't use exim as a dialup - perhaps
someone who would find this useful could contribute a patch -
I understand that Phil has no objection to programming contributions.

Dr. Andrew C. Aitchison        Computer Officer, DPMMS, Cambridge
A.C.Aitchison@???    http://www.dpmms.cam.ac.uk/~werdna




--
*** Exim information can be found at http://www.exim.org/ ***