On 20/10/2022 22:07, Ian Kelling via Exim-users wrote:
> I've had yahoo return a temporary error which says roughly "don't
> email any of our mx hosts for a while".
As Gedalya says, it makes a difference whether you're getting
an error applying to the message or one applying to the host
you're trying to send to. Basically, in response to RCPT or DATA
versus connect, EHLO or MAIL. What do you get logged?
For host errors, a retry rule can use the host name in the "domain"
first field. That, combined with a specific for the second field,
may well do what you need.
For message error responses we'd have to get much more creative;
the retry-rule facilities don't really help (worth a wishlist-level
bug, if you care:
https://bugs.exim.org/ ).
I'd start (assume you run a recent Exim release) by diverting
messages destined for yahoo, in ACL, to a specific named queue -
and have a separate runner for that queue with a slow tick.
I'm unclear if you want *all* messages to yahoo to get held
after one gets this specific error, or if they can be treated
independently. The former is more tricky.
> So,
> I'm setting up 2 exim spools, and 2 daemons, where 1 does fast queue
> runners that send to a second exim daemon which tries an immediately
> delivery but has a different spool, which does the default 30 minute
> queue runners.
You've re-invented Exim's named queues, with the extra hassle of
the re-send.
--
Cheers,
Jeremy
