I have a dial-up connection and followed the FAQ's suggestion (Q1404)
and set up a queue_only_file to cause exim to queue all files until my
connection was up. I'm using a pretty basic smarthost configuration.
After I originally set this up (in version 4.1.2), it worked great. I
had my ip-up script set to start the exim daemon (exim -bd), remove the
queue_only_file, and do a queue run (exim -qqf) about a minute after
the connection was started (to give fetchmail a little time to grab the
mail from my ISP first). My job involves sending out many reports by
email, so I could queue these, and they would all be sent automatically
after the connection was started. My ip-down script killed the daemon
and recreated the queue_only_file.
Recently I created a startup script to just start the exim daemon when
my computer starts up; seemed a bit pointless starting and killing it
all the time. But now when I send emails when not connected, I keep
getting a "host lookup did not complete (DNS timeout?)" defer message
for my smarhost from exim. However if I send an email after connecting
(ie after the queue_only_file is deleted, but before the queue runner
starts), I get the "SMTP delivery explicitly queued" defer message,
which is what I would expect all the time, since the queue_only_file is
in place when I'm not connected. Isn't this behavior backwards?
What am I missing here? Seems like exim should not be doing anything
with these mails except queuing them when the queue_only_file exists.
Is there another configuration option I need to set to keep exim from
attempting to resolve the smarthost when not connected?
