On Fri, 18 Oct 1996, Neal Becker wrote:
> I think I understand what's happening. When more than some amount
> (64k?) is attempted to be delivered to our ccmail/smtp gateway, it
> just rudely drops the connection. This causes the socket to give the
> SIGPIPE.
>
> Unfortunately, this then causes all subsequent mail to get stuck
> waiting for retries!
>
> This is a major problem for me. I wonder what's the best fix?
The *best* fix is clearly to get the gateway mended. :-))
This is the same problem as the user who is near quota - large messages
fail but small ones can be delivered. It is a particular example of
"some particular messages cause temporary delivery failure, while others
can get through".
Exim deliberately processes its queue in "random" order so that messages
of this kind are not necessarily processed before others all the time.
However, that doesn't really help if you have lots of messages piling
up.
The only think I can think of is setting a short retry time for the host
that is broken, in the hope that there will be few enough large messages
for it to get the short ones most of the time. Otherwise you have to do
a lot of nursing by hand.
--
Philip Hazel University Computing Service,
ph10@??? New Museums Site, Cambridge CB2 3QG,
P.Hazel@??? England. Phone: +44 1223 334714
