Bouncing individual email addresses

[Antony Stone]

On Wednesday 27 August 2003 4:17 pm, Rick G wrote:

> On Fri, 8 Aug 2003 17:20:34 +0100, Antony Stone wrote:
> > If you want to avoid the MS processing overhead for these users, why not
> > set rules for "Spam Checks = " so that those users are listed as "no" and
> > not bother to do any spam checking for them at all?   Then just let the
> > relay machine deliver them as quickly as possible to the system which
> > *can* do proper address mapping to either dump the email to /dev/null or
> > else return "no such user".

> Your response on this thread caught my attention. We've just installed MS
> and in tailing the maillog, I noticed the following:
>
> We direct our "catchall" account to /dev/null for exactly the reasons you
> mention. Given our user accounts / aliases are pretty tightly controlled,
> there is no need to bounce back invalid accounts or non-existent aliases.
>
> However, I noticed in the maillog that MS still evaluates these inbound
> messages for Spam/Virus' (and sends a warning notice to the Postmaster)
> before dumping them into /dev/null. Is there anyway to reverse this
> behavior? The only thought I had was to list all valid users / aliases with
> Spam Checks = yes and reverse the default, but that could be a huge effort.
>
> Any ideas? Thx - Rick

Depending on the MTA you are using, is there any easy(ish) way you could use
the list of valid users (email address to mailbox mapping table, or the list
of valid mailboxes perhaps?) to create a 'front-end' valid user list (this
would be virtusertable in sendmail, don't know about others) so that only
mail for valid addresses is accepted, all else is > /dev/null before it gets
near MailScanner?

I think it's better to do this sort of thing at the front-end MTA instead of
inside MailScanner, and presumably you already have some sort of
user-management tool for your MTA to take care of what sounds like a large
number of users?

The sooner you can drop/reject/whatever email to non-addresses the better.

Regards,

Antony.

--

There are two possible outcomes.

If the result confirms the hypothesis, then you've made a measurement.
If the result is contrary to the hypothesis, then you've made a discovery.

 - Enrico Fermi