On 22 June 2002, Adam Henry said:
> Gettings the "defer (-44): retry time not reached for any host" error
> in my logs since about 10pm last night. 10 hours after the first
> log message, mail is finally delivered in one surge. Am I wrong to
> think that the retry rules are in charge of how frequently mail in the
> queue due to this error is sent out again?
I have noticed similar behaviour. I dismissed it at the time, under the
impression that retry rules apply to whole *domains* rather than
individual *messages*. Eg. if there are 50 messages waiting to go to
example.com, Exim won't retry every single message every 15 min (or
whatever), but rather some random message for that domain will be tried
every 15 min. Philip, is that reasoning correct?
Oh wait: if the remote host was only down for five minutes, this doesn't
work. It explains why each message isn't retried every 15 min if the
remote host is down for the whole time, but it doesn't explain why
messages for example.com will stay on the queue after example.com comes
back up.
Greg
