Hello,

We are having the following problem, which I imagine affects a lot of
people:
All machines in this setup are Sendmail 8.13.7.
We have several servers that are the MX record for many domains. These
machines only do spam filtering with Spamassassin, virus filtering with
Clam, and then forward filtered messages to a large internal server.
They do this by the Smarthost parameter in sendmail.cf.

The problem is that we are not able to bounce messages during the smtp
transaction, but are having to reject them when the internal large
server gets the messages. This, of course, is causing us all sorts of
trouble. We are getting blocked by many ISPs and our spam filters are
doing more work than they should.

All of our user accounts are virtual.
How can our spam filters know which users are valid and which ones are
not, and then forward filtered messages to the internal server?

Any ideas or direction will be greatly appreciated.

Arie Kachler

In comp.mail.sendmail ksimports@gmail.com:
> Hello,

> We are having the following problem, which I imagine affects a lot of
> people:
> All machines in this setup are Sendmail 8.13.7.
> We have several servers that are the MX record for many domains. These
> machines only do spam filtering with Spamassassin, virus filtering with
> Clam, and then forward filtered messages to a large internal server.
> They do this by the Smarthost parameter in sendmail.cf.

> The problem is that we are not able to bounce messages during the smtp
> transaction, but are having to reject them when the internal large
> server gets the messages. This, of course, is causing us all sorts of

That's a real problem and is likely to never get you out of
spammer lists. Rejecting as soon as possible should be your goal.

Greylisting still works pretty good, I'd look into it. IMHO we
could get rid of spam if running adaptive teergrubing on just
enough valid MTAs.

> trouble. We are getting blocked by many ISPs and our spam filters are
> doing more work than they should.

> All of our user accounts are virtual.
> How can our spam filters know which users are valid and which ones are
> not, and then forward filtered messages to the internal server?

AFAIK Ldap routing could be used to accomplish a receiver verify,
if one of your internal server can provide the info via ldap
(many can):

http://www.sendmail.org/m4/ldap_routing.html

Good luck

--
Michael Heiming (X-PGP-Sig > GPG-Key ID: EDD27B94)
mail: echo zvpunry@urvzvat.qr | perl -pe 'y/a-z/n-za-m/'
#bofh excuse 266: All of the packets are empty.

ksimports@gmail.com wrote:
> Hello,


> How can our spam filters know which users are valid and which ones are
> not, and then forward filtered messages to the internal server?

This last came up here just last week.

http://groups.google.com/group/comp....201eef96fe73ef