Mike Jackson wrote:
> Sorry for something that's off-topic, but this is the only list I
> read  :)

Please don't hijack threads, when you ask a question, start a new message,
if you reply to an old message and change the subject, those of us who read
this group in thread view might miss your message entirely.

Your mail reader (OE) places headers which aid in the threaded view of mail.
Your message had:
References: <[EMAIL PROTECTED]>

Which places it in the thread titled: "Huh?!?!"

Have you ever used newsgroups before and seen how those messages are grouped
by conversation?  Some of us read our e-mail this way and thread hijacking
is a pain!

Now to help with your question,
Setting up a secondary MX server means having a second box, not creating a
second instance on the same server.  If you have a hardware failure or
someone unplugs the ethernet by mistake, you will not receive any mail.
Many companies who really care about e-mail will even setup a second MX
server off site as to aid in case of T1's going down, power loss, floods,
other natural disasters, you name it!  Ever hear the saying "Shit happens"?
It really does and if you are around long enough, I'm sure you'll see your
share.

If you have a low mail volume, find an old 486, install Linux or some BSD
and you'll be good to go.

It's probably a bad idea to automatically feed anything into sa-learn.
If I was a spammer trying to target you, I could keep sending you the same
message, changing little bits which get caught by your server until I am
able to make your server autolearn my message as ham.  This might not be
possible unless you block during smtp session, but it shows an example of
how auto-learning can come back to bite you.  Or they might pass their mail
through their own copy of SA and ensure that it's not spammy before they
send it.  (Many do this already, but thanks to fast work from SARE, this
isn't such a problem as it was before.)

As for auto-learning all outgoing mail, I found this to be a bad idea too,
just wait until someone gets infected with a virus / spam relayer and starts
using your smtp server to relay it's spam.  Once you auto-learn all that
spam as ham, your bayes will be toast.

Another possible use for this would be to forward mail to this special
account for learning, this can be a problem too depending on if your mail
client adds or modifies the message when forwarding, you migth not end up
with the bayes data you would if you learned the message directly.

Frederic Tarasevicius
Internet Information Services, Inc.
http://www.i-is.com/
810-794-4400
mailto:[EMAIL PROTECTED]

Reply via email to