On 10 Nov 2004 at 17:02, Rob wrote about
"Re: [exim] unrouteable address from":
|...
| The only error is ** Unrouteable address.
Meaning all of your routers declined (as opposed to an ACL
rejection). That begs the question why the message was not rejected
by a verify=recipient test in the RCPT ACL.
|...
| Here is a slightly obfuscated snip from the log,
|...
| |--- 2 --- |
|
| 2004-11-10 10:01:42 1CRtyf-0007o6-Vx SA: Action: Not running SA because
| SAEximRunCond expanded to false (Message-Id: 1CRtyf-0007o6-Vx). From
| <mail-sender@???> (host=internalhost.ourdomain.com
| [65.217.xyz.xyz]) for our-user@???
|
| 2004-11-10 10:01:42 1CRtyf-0007o6-Vx <= mail-sender@???
| H=internalhost.ourdomain.com [65.217.xyz.xyz]:40603
| I=[65.217.abc.abc]:25 U=Debian-exim P=esmtp X=TLS-1.0:RSA_ARCFOUR_SHA:16
| S=2697 id=41922D48.4020003@??? T="test" from
| <mail-sender@???> for our-user@???
|
| 2004-11-10 10:01:42 1CRtyf-0007o6-Vx ** our-user@???
| F=<mail-sender@???>: Unrouteable address
Seems clear enough. The message was not rejected by any ACL, but
none of your routers were triggered by it. This is a Bad Thing,
unroutable messages should be rejected in the RCPT ACL. But it
should be easy to debug via exim -bhc (as previously suggested) if
you can't spot it simply by eyeballing the router you *expect* to
handle this address (dnslookup, in the sample config).
Did you run the "exim -bhc <internal-server-IP-address>" session on
the external box? Running it on the internal box won't tell you
anything.
- Fred
