> -----Original Message-----
> Subject: Re: [AMaViS-user] AWL score not being added correctly?
> |
> | So, I agree we have a bug / feature request here.
> |
> | Mark M. - Any chance of updating the scores output on non-passed
> | messages such that they reflect the true value used in
> determining the
> | message's disposition?
>
> The sender_score is a per-recipient thing. The locations
> where only per-message information is reported do not
> necessarily represent things accurately, e.g. if one
> recipient blacklists a mail and the other not, is the message
> blacklisted or not? Similarly for the final score, which is
> a sum of per-message score as reported by SA, plus the
> per-recipient score_sender boost.
>
> The main log entry with 2.3.2 now reports hits as an explicit
> sum of SA score and sender score to make it more obvious.
>
> The quarantined message only contains one set of X-Spam and
> other inserted headers, even when the message was addressed
> to more than one recipient and their settings (e.g. kill
> level) were different. To reflect the truth, there should
> either be several quarantined copies, one for each recipient,
> or there should be perhaps multiple instances of X-Spam*
> headers, one set for each recipient. At the moment this is
> not so, the information is 'somehow'
> aggregated into a per-message set of headers, with a
> necessary loss of information, such a per-recipient
> score-sender boosts.
>
> The main log entry and the passed message are accurate
> though, and so is the SQL report/log, which has a spam score
> sum in per-recipient table.
>
> The problem is noted in TODO, but I don't know what the best
> solution would be.
Mark,
As usual, thanks for your excellent response. This makes perfect sense. A
difficult situation indeed. I don't think it would be prudent to include
multiple X-Spam headers, as recipient A needs to know nothing about
recipient B's config.
For future reference to others, below are the two log entries for the
message in question: the first shows the per-message with the raw SA score
(1.537), and the second shows the per-recipient score, which includes the
score_sender boost (1.537+5). Until your explanation, I had only really
paid attention to the first entry, but see the errors in my ways.
Jul 14 18:09:55 glacier amavis[450]: (00450-08) SPAM,
<[EMAIL PROTECTED]> -> <[EMAIL PROTECTED]>,
No, hits=1.537 tag=2 tag2=6.31 kill=6.31
tests=[AWL=0.612, BAYES_40=-1.096, HTML_80_90=0.146,
HTML_FONT_BIG=0.142, HTML_MESSAGE=0.001, MSGID_FROM_MTA_ID=1.723,
SPF_HELO_PASS=-0.001, SPF_PASS=-0.001, URI_REDIRECTOR=0.011],
autolearn=no, quarantine iFN64e4g-YL1 (spam-quarantine)
Jul 14 18:09:55 glacier amavis[450]: (00450-08) Blocked SPAM,
[64.40.98.25] [64.40.98.25] <[EMAIL PROTECTED]> ->
<[EMAIL PROTECTED]>,
quarantine: spam-iFN64e4g-YL1.gz,
Message-ID: <[EMAIL PROTECTED]>,
mail_id: iFN64e4g-YL1, Hits: 1.537+5 ...
Thanks,
Mike
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
AMaViS-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/amavis-user
AMaViS-FAQ:http://www.amavis.org/amavis-faq.php3
AMaViS-HowTos:http://www.amavis.org/howto/
