Thanks for the response --

I can't 100% understand the docs, but AFAICT it seems to check the
rcvd from last handover from an untrusted network.  so am i right in
thinking i have to add all of the servers for all the email forwarding
services i use to my trusted network for this to work?  or does
whitelist_from_rcvd check the originating server where the mail came
from?

in more detail: my problem is that i'm getting (not very much) spam
from an email fowarding service i have (from an alumni network.)  e.g.

[EMAIL PROTECTED] forwards to [EMAIL PROTECTED]

some spammer is sending email to [EMAIL PROTECTED], with a From: saying
it's from [EMAIL PROTECTED]  this is getting forwarded to
[EMAIL PROTECTED]  it is being received by the myemail.com mail servers
from the college.edu mailservers, so checking the from matches the
rcvd will pass, i think.  i need to check that the from matches the
rcvd farther down the line, e.g. the original message to
[EMAIL PROTECTED]  do i handle this by making the college.edu
mailserver trusted?

right now i'm blacklisting myself to combat this problem.  not the
most ideal solution.

B

On Thu, 26 Feb 2004, Dan DeVoe wrote:

> Ben,
>
> You may want to try using the "whitelist_from_rcvd" function rather than
> just bare "whitelist_from". I got confused on this issue myself, but 2.60
> added that new function specifically to stop this kind of abuse, Using
> whitelist_from_rcvd will check the Received: headers as well as the From:
> address.
>
> You can refer to the docs here:
>
> http://www.spamassassin.org/doc/Mail_SpamAssassin_Conf.html#whitelist%20and%20blacklist%20options
>
> for the exact syntax and functionality of whitelist_from_rcvd, which is
> not identical to that of whitelist_from.
>
> If we had a FAQ, I would suggest this for one of the questions at this
> point. =)
>
> --
>  .''`.     Daniel DeVoe <[EMAIL PROTECTED]>
> : :'  :    http://www.netset.com/~ddevoe
> `. `'`
>   `-  Debian - when you have better things to do than fix a system
>
> On Wed, 25 Feb 2004, Ben wrote:
>
> > Date: Wed, 25 Feb 2004 00:03:58 -0500 (EST)
> > From: Ben <[EMAIL PROTECTED]>
> > To: Keith C. Ivey <[EMAIL PROTECTED]>
> > Cc: [EMAIL PROTECTED]
> > Subject: Re: abusing the whitelist
> >
> > thanks for the response and the advice.  unfortunately it's a mail
> > forwarding system that i'm part of that i'm trying to whitelist
> > (alumni email forwarding) so the mail is not actually local.  i guess
> > i'll think about this more deeply.
> >
> > B
> >
> > On Tue, 24 Feb 2004, Keith C. Ivey wrote:
> >
> > > Ben <[EMAIL PROTECTED]> wrote:
> > >
> > > > the only reason why this got through was they put one of my email
> > > > address in the From, in quotes.  i'm no RFC822 or whatever expert, but
> > > > that seems like it shouldn't get tagged by the USER_IN_WHITELIST test.
> > >
> > > Perhaps, but even if it didn't match addresses that weren't
> > > really addresses, the spammer could just as well have put your
> > > address in the "From:" line without quotes.  In fact, I'd say
> > > that's a lot more common than the quoted version you're
> > > reporting.  Whitelisting your own address, or any addresses at
> > > your domain, is generally a bad idea, since you will end up
> > > whitelisting a lot of spam.
> > >
> > > One alternative is to have mail that's actually from local
> > > users not go through SpamAssassin at all.  The details of how
> > > to do that depend on your mail system.
> > >
> > > --
> > > Keith C. Ivey <[EMAIL PROTECTED]>
> > > Washington, DC
> > >
> > >
> >
> >
>



Reply via email to