MailScanner spam filtering and Thunderbird undesirable mails

[Scott Silva]

Sam Przyswa spake the following on 12/12/2006 9:33 AM:
> Hi,
> 
> I'm not sure that my MailScanner/Spamassassin is well configured, on a
> machine the command "sa-learn --dump all" return that:
> 
> <18:03:19>root at mjc-idf ~ >sa-learn --dump all
> 0.000          0          3          0  non-token data: bayes db version
> 0.000          0          0          0  non-token data: nspam
> 0.000          0          2          0  non-token data: nham
> 0.000          0        502          0  non-token data: ntokens
> 0.000          0 1158007683          0  non-token data: oldest atime
> 0.000          0 1158007690          0  non-token data: newest atime
> 0.000          0          0          0  non-token data: last journal
> sync atime
> 0.000          0          0          0  non-token data: last expiry atime
> 0.000          0          0          0  non-token data: last expire
> atime delta
> 0.000          0          0          0  non-token data: last expire
> reduction count
> 0.500          0          1 1158007683  5a2a12092d
> 0.500          0          2 1158007690  87450837a1
> 0.500          0          1 1158007683  8492dae09c
> ...
> 
> With:
> SpamAssassin version 3.1.4
>  running on Perl version 5.8.8
> 
> On that machine the spam filtering seems working fine.
> 
> On other machine the command "sa-learn --dump all" give me only that:
> 
> root at soleil:/tmp# root at soleil:/etc/MailScanner# sa-learn --dump all
> 0.000          0          3          0  non-token data: bayes db version
> 0.000          0          0          0  non-token data: nspam
> 0.000          0          0          0  non-token data: nham
> 0.000          0          0          0  non-token data: ntokens
> 0.000          0          0          0  non-token data: oldest atime
> 0.000          0          0          0  non-token data: newest atime
> 0.000          0          0          0  non-token data: last journal
> sync atime
> 0.000          0          0          0  non-token data: last expiry atime
> 0.000          0          0          0  non-token data: last expire
> atime delta
> 0.000          0          0          0  non-token data: last expire
> reduction count
> 
> With:
> SpamAssassin version 3.1.0
>  running on Perl version 5.8.7
> 
> With this machine my Mozilla-thunderbird find lot of undesirable mails,
> does Thunderbird is more effective than MailScanner/Spamassassin or my
> setup is bad configured ?
> 
> Thanks in advance for your help.
> 
> Sam.
> 
> 
> 
You have almost no bayes data. Here is a --dump magic from one of my lesser
hit servers;

0.000          0          3          0  non-token data: bayes db version
0.000          0     230194          0  non-token data: nspam
0.000          0     148815          0  non-token data: nham
0.000          0     193563          0  non-token data: ntokens
0.000          0 1165425289          0  non-token data: oldest atime
0.000          0 1165946146          0  non-token data: newest atime
0.000          0 1165945429          0  non-token data: last journal sync atime
0.000          0 1165864479          0  non-token data: last expiry atime
0.000          0          0          0  non-token data: last expire atime delta
0.000          0          0          0  non-token data: last expire reduction
count

As you can see, the nham and nspam counts are in the six figure range.
Could you be running spamassassin as another user, and maybe you dumped the
wrong database?

-- 

MailScanner is like deodorant...
You hope everybody uses it, and
you notice quickly if they don't!!!!