RE: [exim] Most Lenient RBL implimentation?

トップ ページ
このメッセージを削除
このメッセージに返信
著者: Herb Martin
日付:  
To: 'Steven B', exim-users
CC: 
題目: RE: [exim] Most Lenient RBL implimentation?
> I've recently been having some problems with SORBS being a little too
> reactive. As an ISP we can't afford to be quite as fascist as a small
> business or office. I apologize if this is not very EXIM
> specific, and
> would welcome any references to better places to ask this question.
>
> Any opinions on which RBL's are the most, well, "laid back"
> -- kind of a
> "only the worst of the worst" super conservative RBL?


My advice is to drop almost all RBL for REJECTING
email and use them STRICTLY for "Greylisting" email.

You can use even the most aggressive RBLs and have
practically a zero false positive rate without doing
any serious whitelist maintenance.

This isn't precisely what you asked but as an ISP
you really have no true idea what mail is tolerable
for your users.

While that argument COULD be made for greylisting the
counter is that greylisting ONLY requires that the
sender use a "real SMTP" server that 1) meets the
normal requirements for retry OR 2) is NOT listed
on major RBLs.

We practically never reject email on a single criteria,
but use RBLs, SpamAssassin, headers checks, etc to
DRIVE greylisting.

Those that pass greylisting may be rejected but ONLY
if they fail on several of the other criteria.

--
Herb Martin

> -----Original Message-----
> From: exim-users-bounces@???
> [mailto:exim-users-bounces@exim.org] On Behalf Of Steven B
> Sent: Monday, March 13, 2006 2:46 PM
> To: exim-users@???
> Subject: [exim] Most Lenient RBL implimentation?
>
>
> I dropped most of the SORBS lists and have the following
> still in place:
>
> sbl-xbl.spamhaus.org
> list.dsbl.org
> smtp.dnsbl.sorbs.net
> relays.ordb.org
> dnsbl.njabl.org
> opm.blitzed.org
> bl.spamcop.net
>
> Any comments appreciated.
>
> Steve
>
>
>
>
>
>
>
>
>
>
> --
> ## List details at http://www.exim.org/mailman/listinfo/exim-users
> ## Exim details at http://www.exim.org/
> ## Please use the Wiki with this list - http://www.exim.org/eximwiki/
>