Mailing List Archive

Declare as spam with learning for failed DMARC/SPF
Hello,

I have Ubuntu 18.04, Postfix 3.3, OpenDMARC 1.3.2, OpenDKIM 2.11,
Spamassassin 3.4.2, and Dovecot 2.2.

I am trying to achieve a configuration where:

If there is a DMARC test and it fails, then send to spam directory.

If there is no DMARC test and there is an SPF test and it fails, send to
spam directory.

If the user moves the email out of the spam directory, Spamassassin
learns that this is not spam, even with DMARC or SPF fail.

I apologize if this is some obvious configuration, but in my reading I'm
not able to discover and searching produces a considerable amount of noise.

I initially tried to achieve this using sieve, but it seems maybe sieve
ignores any of the learning or I didn't discover how to create the rule.

Thank you,

Paul
Re: Declare as spam with learning for failed DMARC/SPF [ In reply to ]
On 14.05.20 05:59, Paul Pace wrote:
>I have Ubuntu 18.04, Postfix 3.3, OpenDMARC 1.3.2, OpenDKIM 2.11,
>Spamassassin 3.4.2, and Dovecot 2.2.
>
>I am trying to achieve a configuration where:
>
>If there is a DMARC test and it fails, then send to spam directory.

>If there is no DMARC test and there is an SPF test and it fails, send
>to spam directory.

SpamAssassin can't send anything anywhere. SpamAssassin only scores e-mails
and decides if they are spam or ham.

I think SpamAssassin doesn't have DMARC plugin but simple google search
answered:

https://notes.sagredo.eu/en/qmail-notes-185/setting-dmarc-filter-in-spamassassin-236.html

you can score various DMARC and SPA rules to mark the mail as spam, and it's
up to your delivery agent to move spam to spam directory,


>If the user moves the email out of the spam directory, Spamassassin
>learns that this is not spam, even with DMARC or SPF fail.

I think dovecot IMAP has plugin to do this.

>I apologize if this is some obvious configuration, but in my reading
>I'm not able to discover and searching produces a considerable amount
>of noise.
>
>I initially tried to achieve this using sieve, but it seems maybe
>sieve ignores any of the learning or I didn't discover how to create
>the rule.

--
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Your mouse has moved. Windows NT will now restart for changes to take
to take effect. [OK]