Stephen Bosch <[EMAIL PROTECTED]> wrote:
> > Stephen F. Bosch <[EMAIL PROTECTED]> wrote:
> > > What about assigning this user a special SMTP port? Can I do this with
> > > qmail? Run an additional SMTP service at a custom port number? That
> > > would bypass such a block.
> >
> > Yes. Run another copy of qmail-smtpd on an arbitrarily chosen high port.
>
> How is this done? I take it there is a switch for it...?
Run qmail-smtpd from tcpserver, and specify a different port number in the
port argument.
> > Also, you might want to use a smarter relay control mechanism than opening
> > up an entire /16 for relaying; you could easily get hit that way. Perhaps
> > an SMTP-after-POP solution, like Bruce Guenter's relay-ctrl package.
> *Everybody* says that =) But I am using IMAP, and as far as I know AT&T
> has the entire 12. class A... unless there are some major spammers at
> AT&T, I should be okay...
There are lots of spammers on all dialup services; I get enough from AT&T
customers to know. This is probably why they've blocked SMTP from their
customers out to the rest of the net.
> I know that sounds really dumb. Has anybody got a better idea? I'm open.
>
> If I knew the exact address ranges of above person's dialup I would be
> able to narrow it down significantly, but I have no real way of
> determining this - and AT&T is unlikely to be very cooperative, given
> their decision to close all internal to external SMTP traffic.
Use an SMTP-after-POP method; Bruce Guenter's relay-ctrl package is excellent
for this. Basically, you make the user do a POP3 check before trying to
send mail; after a successful authentication with POP3, that IP address is
allowed to relay through your SMTP daemon for (default) 15 minutes.
Have a look at http://www.em.ca/~bruceg/relay-ctrl/ IIRC.
Charles
--
--------------------------------------------------------------
Charles Cazabon <[EMAIL PROTECTED]>
QCC Communications Corporation Saskatoon, SK
My opinions do not necessarily represent those of my employer.
--------------------------------------------------------------
