On Tue, 30 Sep 2014, terrygalant.li...@fastest.cc wrote:
Hi John
On Tue, Sep 30, 2014, at 10:11 AM, John Hardin wrote:
How are you training your Bayes database?
By submitting the emails to a spam-training email address @ my Zimbra server,
http://wiki.zimbra.com/wiki/Zimbra_MTA#Anti-Spam_Training_Filters
Okay, since it's a packaged solution most of my concerns don't apply. I
would assume Zimbra would internally be Doing the Right Thing.
How much have you trained it? It requires a certain minimum amount of both
spam *and* ham before it starts evaluating messages.
I've been running this server for a couple of years.
What is the output of sa-learn --dump magic ?
/opt/zimbra/libexec/sa-learn --dbpath /opt/zimbra/data/amavisd/.spamassassin
--dump magic
0.000 0 3 0 non-token data: bayes db version
0.000 0 1150 0 non-token data: nspam
0.000 0 19454 0 non-token data: nham
You might want to verify the "nspam" and "nham" numbers change as you feed
in messages for training.
One common error is to train Bayes as a different user than SA (spamd,
amavisd) is running under. What user is spamd running under, and what user
are you running sa-learn to train under?
Here I get a little confused. Likely the source of at least one of my problems
:-/
Iiuc, SA is managed under/by amavisd. It runs as the 'zimbra' user.
Did you run the above as user zimbra?
Because SpamAssassin is run by amavisd-new I guess you have to run the
sa-learn utility as the amavis user, which in this case is 'zimbra'.
Right.
The example you gave, one of those background check spams, should have
enough content to trigger a high bayes score.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhar...@impsec.org FALaholic #11174 pgpk -a jhar...@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
Maxim VI: If violence wasn’t your last resort, you failed to resort
to enough of it.
-----------------------------------------------------------------------
4 days until the 10th anniversary of SpaceshipOne winning the X-prize
