On Tue, 18 Jun 2002, Philip Hazel wrote:
> I don't quite understand this. Why weren't the queue runners picking
> this up?
oh, should have my my point more clear. that's because i have exiscan
running on the box, and it operates as exim being in queue_only mode,
and exiscan picks messages up from the queue, scans them, and
reinjects and runs them.
so, the behaviour is exactly as it should be, at least with regards to
exim.
as far as i understand exiscan, it picks messages, scans, then tries to
make exim to deliver it. if it fails, the message stays on the queue,
exiscan clears its msgid cache (of scanned messages) every now and
then (1hours with my setup), so it will try to re-run it every now
and then. but, if it fails, the message is just staying there...
basically, the question is, how do others circumvent this "loop",
short of regularly looking at the queue and running it manually ? i
could of course cook up some q&d solution to look at the queue, pick
messages which are there for longer than X, and -qff -d them (which as
i'm observing does obey the retry rules), i'm just trying to get a
sense of how others would do it.
sorry if i was confusing at the first place. (no "if" actually. i was
;)
