On Tue, 17 Nov 1998, Clay Porter wrote:
> Sorry for the poorly worded question... I understand that if a message
> cannot be delivered then it will be queued and exim will retry delivery
> according to the rules. However, when I re-started the MTA on the receiving
> host I did not expect *subsequent* messages to be deferred for retry; I
> expected those messages to go through right away.
No, it doesn't work like that. Exim keeps retry information *per host*
(strictly, per IP address). If a host has been down for a long time, it
won't try that host, for *any* message, until its retry time comes. This
has two benefits:
. While the host is down, your host doesn't keep wasting resources
trying it every time it has a message for it.
. When the host comes back, other mailers out in the world won't all
suddenly start to hammer it at once, because their retry times are
likely to be different. At least, that's what would happen if they all
used the same strategy. In practice, many of them seem to try much
more frequently.
As soon as one message has been successfully delivered to the host, the
next queue run should deliver the rest. Therefore, if you know a host
has come alive again and want to speed things up, all you need to do is
to prod one message with the -M option or via eximon.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
