On Mon, 20 Sep 1999, David Harris wrote:
> I'm thinking of deploying RBL to try to cut down on spam, but before I did that
> I wanted to poke around and see how effective it might be. So, I gathered up
> some spam messages that I had received and looked up the mailserver's ipaddr in
> RBL using rbl.maps.vix.com and rbl.dorkslayers.com, and not one host was
> rejected from either RBL site. Even though I could see the messages looked like
> they were going trough an open relay.
>
> How good is this whole RBL thing anyway?
It's not terribly pro-active, and they won't RBL anyone if you just
forward them the headers from a spam, even if they have verified that it
came from an open relay.
You have to demonstrate that the server admin(s) of said relay was
unresponsive or uncooperative in taking steps to shut down the open relay
first.
I can certainly see why they do things this way, but it definitely limits
the RBL's effectiveness as a spam filtering mechanism. OTOH, it's been
outstanding in terms of getting lethargic providers (big and small) to
crack down on spamming customers.
It still appears to catch a significant number of spams (just not enough
to make an impact on my system or mailbox), and IMHO, is worth
implementing, but you should also employ the MAPS DUL, which appears to
catch almost all of the spam that comes directly from dialup accounts (not
through a relay). It also appears that the spammers have caught on to
that and have reverted to using open relays, and unfortunately, there's
still an abundance of them, and more coming on line all the time.
I've been toying with using ORBS (I already forward open relays to them),
but am reluctant for various reasons.
