Re: [exim] Retry time not reached for any Host

On 2008-01-03 at 16:02 -0500, Joel Alexander wrote:
> Can someone please help me? I have several messages that get thrown to
> my fallback_hosts for no known reason. Is there a way I can fix this?

> 2. Immediately the message in the log shows:
> LOG:
> 2008-01-03 11:20:29 1JASo1-00042K-9p == cebersole@???
> R=dnslookup T=remote_smtp defer (-53): retry time not reached for any
> host
>
> 3. Then it is pushed to my fallback host. Why?

Exim caches the availability of remote hosts. This means that when a
popular remote site goes down, Exim doesn't kill your system repeatedly
trying to deliver whilst it's down.

Use "exinext" to find out what error was seen and when the next retry
would be seen. See The Exim Specification:
50.10 Finding individual retry times (exinext)
(section number accurate with current Exim).

If you're automatically passing all mail to a fallback host if it's not
deliverable immediately then you probably don't want to refuse initial
delivery based on a long-term cache; a short-lived cache so that a run
of 20 mails in a short space of time won't keep hitting the limit is
probably still useful though.

# this in the main configuration section:
retry_data_expire = 1h

The default is 7 days, which makes sense for a standalone server.

-Phil