exm user wrote:
>
> Dear all,
>
> As described in the subject, my exim server on and off can't send mail
> to some external email address, for example, if sending 10 emails to
> external address, there will be 1 or 2 mail bounced back with error
> "Unrouteable address".
>
As a realy realy realy simple question, are you absolutely sure that the
addresses that are 'Unrouteable' are proper? as in absolutely positively no
difference between those and the ones that route properly?
As a second question, are there any other errors in the exim logs that
tell you when and where those addresses are being routed to, and is it the
same as when routing the delivered messages?
>
> my /etc/resolv.conf is pointing to correct name server, and i can
> actually get the external server mx record, and even telnet inside.
> while i tried to manually telnet or exim -bh testing from my mailserver,
> it alwasy working and can get pass the ACL without any error, i just
> can't figure out why that problem keep rising.
>
> and one thing i don't understand is that i already set the
> ignore-bounce-error for 1 day, timeout-frozen-after for 3 days, and
> retry configuration of entire queue for 4 days, why the unroutable
> address won't be kept in the queue if the external host failed, instead
> it straight away return back to me..
>
If the message is not routable that is the final delivery attempt, why keep
the message on the queue when it cannot be routed to a destination.
This is a permanant failure, not a temporary one, your exim server has no
clue how to deliver these messages.
>
> is there any exim hidden dns setting that will randomly fail ? or
> there's some error in my retry config?
> i am using exim 4.05, with amavis-snapshot200203 and f-prot.
>
> following is my retry config:
>
> * * F,2h,15m; G,16h,1h,1.5; F,4d,6h
> ignore_bounce_errors_after = 1d
> timeout_frozen_after = 3d
>
>
> please advice on anything i need to check, or anything i could do, need
> help asap, thanks everyone!
>
I would advise that there is nothing wrong in the retry, however, depending
on HOW you are scanning your messages (i.e. during routing and not during
SMTP DATA) the 'deliver to?' could be getting munged to something that is
no longer routable, even though it may have been before the scans...
That is just conjecture as you do not mention HOW or when you are scanning
the messages.
>
> exim faithful user
> --
Same here :)
--
--EAL--
