On Tue, 11 Jan 2000, Nigel Metheringham wrote:
> exim@??? said:
> > I have seen this exact behaviour just last weekend. An address on a
> > mailing list was failing - not with DNS but, because the SMTP server
> > in question was down. However the site in question had 6 SMTP servers
> > and, exim would sit and try each server and fail, each time a message
> > to the list was posted (tis a fairly busy list).
>
> This should be caught by the retry rules surely unless you have very
> short retry times since the first (few) messages would cause a block,
> but then hint entries would be added to the retry db and prevent
> further attempts until the retry time had passed.
Aha! This could explain the originally reported behaviour: "Most
messages get delivered fine, but every now and again one stalls" (I
paraphrase). No doubt the "every now and again" is when the retry time
for the broken host or hosts arrives.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
