Spamassassin autolearn=no



  • Hello,

    i’am using postfix, amavis, spamassassin and clamav.

    Amavis triggers spamassassin and clamav.

    Every mail which arrives in kopano includes this entry in the mail meader:
    autolearn=no autolearn_force=no

    Perhaps this is a normal behaviour, but i thought when i use bayes then it should autolearn, or iam false?

    To explain spamassassin which mails are spam iam using inotify-spamlearn and kopano-spamd.

    This is my /etc/spamassassin/local.cf:

    # This is the right place to customize your installation of SpamAssassin.
    #
    # See 'perldoc Mail::SpamAssassin::Conf' for details of what can be
    # tweaked.
    #
    # Only a small subset of options are listed below
    #
    ###########################################################################
    
    #   Add *****SPAM***** to the Subject header of spam e-mails
    #
    # rewrite_header Subject *****SPAM*****
    
    
    #   Save spam messages as a message/rfc822 MIME attachment instead of
    #   modifying the original message (0: off, 2: use text/plain instead)
    #
    # report_safe 1
    
    
    #   Set which networks or hosts are considered 'trusted' by your mail
    #   server (i.e. not spammers)
    #
    # trusted_networks 212.17.35.
    
    
    #   Set file-locking method (flock is not safe over NFS, but is faster)
    #
    # lock_method flock
    
    
    #   Set the threshold at which a message is considered spam (default: 5.0)
    #
    # required_score 5.0
    
    
    #   Use Bayesian classifier (default: 1)
    #
    use_bayes 1
    
    
    #   Bayesian classifier auto-learning (default: 1)
    #
    bayes_auto_learn 1
    
    
    #   Set headers which may provide inappropriate cues to the Bayesian
    #   classifier
    #
    bayes_ignore_header X-Bogosity
    bayes_ignore_header X-Spam-Flag
    bayes_ignore_header X-Spam-Status
    
    
    #   Whether to decode non- UTF-8 and non-ASCII textual parts and recode
    #   them to UTF-8 before the text is given over to rules processing.
    #
    # normalize_charset 1
    
    #   Some shortcircuiting, if the plugin is enabled
    # 
    ifplugin Mail::SpamAssassin::Plugin::Shortcircuit
    #
    #   default: strongly-whitelisted mails are *really* whitelisted now, if the
    #   shortcircuiting plugin is active, causing early exit to save CPU load.
    #   Uncomment to turn this on
    #
    # shortcircuit USER_IN_WHITELIST       on
    # shortcircuit USER_IN_DEF_WHITELIST   on
    # shortcircuit USER_IN_ALL_SPAM_TO     on
    # shortcircuit SUBJECT_IN_WHITELIST    on
    
    #   the opposite; blacklisted mails can also save CPU
    #
    # shortcircuit USER_IN_BLACKLIST       on
    # shortcircuit USER_IN_BLACKLIST_TO    on
    # shortcircuit SUBJECT_IN_BLACKLIST    on
    
    #   if you have taken the time to correctly specify your "trusted_networks",
    #   this is another good way to save CPU
    #
    # shortcircuit ALL_TRUSTED             on
    
    #   and a well-trained bayes DB can save running rules, too
    #
    # shortcircuit BAYES_99                spam
    # shortcircuit BAYES_00                ham
    
    endif # Mail::SpamAssassin::Plugin::Shortcircuit
    
    ifplugin Mail::SpamAssassin::Plugin::RelayCountry
    add_header all Relay-Country _RELAYCOUNTRY_
    header RELAYCOUNTRY_BAD X-Relay-Countries =~ /(CN|RU|UA|RO|VN)/
    describe RELAYCOUNTRY_BAD Relayed through spammy country at some point
    score RELAYCOUNTRY_BAD 2.0
     
    header RELAYCOUNTRY_GOOD X-Relay-Countries =~ /^(DE|AT|CH)/
    describe RELAYCOUNTRY_GOOD First untrusted GW is DE, AT or CH
    score RELAYCOUNTRY_GOOD -0.5
    endif # Mail::SpamAssassin::Plugin::RelayCountry
    
    score RCVD_IN_BL_SPAMCOP_NET 0 5.246 0 5.347
    score RCVD_IN_BRBL_LASTEXT 0 5.246 0 5.347
    score URIBL_BLACK 0 5.7 0 5.7
    score URIBL_WS_SURBL 0 2.659 0 2.608
    score URIBL_MW_SURBL 0 2.263 0 2.263
    score URIBL_CR_SURBL 0 2.263 0 2.263
    score URIBL_GREY 0 2.084 0 1.424
    score URIBL_DBL_SPAM    0 4.5 0 4.5
    score URIBL_DBL_PHISH   0 4.5 0 4.5
    score URIBL_DBL_MALWARE 0 4.5 0 4.5
    score URIBL_DBL_BOTNETCC 0 4.5 0 4.5
    score URIBL_DBL_ABUSE_SPAM 0 4.0 0 4.0
    score URIBL_DBL_ABUSE_PHISH 0 4.5 0 4.5
    score URIBL_DBL_ABUSE_MALW  0 4.5 0 4.5
    score URIBL_DBL_ABUSE_BOTCC 0 4.5 0 4.5
    

    I gave already more then 500 spam mails to learn to spamassassin.

    I think some counts should be higher as zerosa-learn --dump magic

    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
    

    But the logs from inotify-spamlearn says it has Learned tokens:

    INFO Processing [Inotify] /var/lib/kopano/spamd/spam/FC102D47DEDD4AEEABE91BBD90F08D0D.eml: Learned tokens from 1 message(s) (1 message(s) examined)
    INFO Removing file: /var/lib/kopano/spamd/spam/FC102D47DEDD4AEEABE91BBD90F08D0D.eml
    INFO Processing [Inotify] /var/lib/kopano/spamd/spam/F507443E09D24102911540197DC2C66B.eml: Learned tokens from 1 message(s) (1 message(s) examined)
    INFO Removing file: /var/lib/kopano/spamd/spam/F507443E09D24102911540197DC2C66B.eml
    INFO Processing [Inotify] /var/lib/kopano/spamd/spam/66113D3562C548219755ADCF12DD61EC.eml: Learned tokens from 1 message(s) (1 message(s) examined)
    INFO Removing file: /var/lib/kopano/spamd/spam/66113D3562C548219755ADCF12DD61EC.eml
    INFO Processing [Inotify] /var/lib/kopano/spamd/spam/D6E90D3DE8724866A21AE80146022E07.eml: Learned tokens from 1 message(s) (1 message(s) examined)
    INFO Removing file: /var/lib/kopano/spamd/spam/D6E90D3DE8724866A21AE80146022E07.eml
    INFO Processing [Inotify] /var/lib/kopano/spamd/spam/1A600DC9E18A40FA8A925837673A65F8.eml: Learned tokens from 1 message(s) (1 message(s) examined)
    INFO Removing file: /var/lib/kopano/spamd/spam/1A600DC9E18A40FA8A925837673A65F8.eml
    


  • This problem is solved.

    i forgot to set the bayes_auto_learn_threshold values.


Log in to reply