bayes stopped working (now working)

[Nick Meverden]

    [ The following text is in the "iso-8859-1" character set. ]
    [ Your display is set for the "US-ASCII" character set.  ]
    [ Some characters may be displayed incorrectly. ]

> Nick
>
> had somehting similar a couple of weeks ago. The bayes side of SA seemed
> to have got corrupted somehow. Everything reported OK, but doing
> anything on the bayes (sa-learn etc) just timed out.
>
> I restored from the previous days backup and everything was fine after
> that....

I noticed I was no longer getting autolearn=not spam in the email headers
so I removed auto-whitelist and auto-whitelist.mutex and restarted
MailScanner and bayes started working again.  Was this a corrupt
auto-whitelist db?


>
> Nick Meverden wrote:
>> MailScanner stopped using bayes today with no errors, the database is no
>> longer auto-learning (ntokens do not increase) and the BAYES_ score
>> rules
>> are no longer being used.  Has anyone seen this before?  Using
>> spamassassin -D --lint returns no bayes errors and if I run mail through
>> it it'll use bayes.  MailScanner.conf and spam.assassin.prefs.conf both
>> are pointing to the correct db path.  And bayes is enabled in
>> spam.assassin.prefs.conf.  Journal sync is setup in MailScanner.conf for
>> daily.  And I setup auto_expire to 0 in spam.assassin.prefs.conf and run
>> a
>> cron job that shuts down MailScanner and manually expires the db and
>> removes any bayes_expire* files if they exist.  The bayes db was not
>> upgrade from 2.6X it was created using SA 3.0.0.  I would like to be
>> able
>> to save the database with out clearing it out.
>>
>>  thanks,
>>
>>      - Nick
>>
>> Here is the version info, output from sa-learn --dump magic, and a lock
>> file that always exist in /tmp, and the contents of my bayes db.
>>
>> Spamassassin 3.0.0
>> MailScanner Version Number = 4.34.8
>> Perl 5.8.5
>> Linux 2.4.26
>>
>> 0.000          0          3          0  non-token data: bayes db version
>> 0.000          0       2141          0  non-token data: nspam
>> 0.000          0       1389          0  non-token data: nham
>> 0.000          0     116161          0  non-token data: ntokens
>> 0.000          0 1098405034          0  non-token data: oldest atime
>> 0.000          0 1099095787          0  non-token data: newest atime
>> 0.000          0          0          0  non-token data: last journal
>> sync
>> atime
>> 0.000          0 1099355126          0  non-token data: last expiry
>> atime
>> 0.000          0     691200          0  non-token data: last expire
>> atime
>> delta
>> 0.000          0      20945          0  non-token data: last expire
>> reduction count
>>
>> /tmp/MS.bayes.rebuild.lock
>> SpamAssassin Bayes database locked for use by MailScanner 2022
>>
>> /var/spool/MailScanner/spamassassin
>>
>> 2592 auto-whitelist           4 bayes.mutex            328 bayes_seen
>>    4 auto-whitelist.mutex     4 bayes_journal         2604 bayes_toks
>>
>> ------------------------ MailScanner list ------------------------
>> To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
>> 'leave mailscanner' in the body of the email.
>> Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
>> the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
>
> **********************************************************************
>
> This email and any files transmitted with it are confidential and
> intended solely for the use of the individual or entity to whom they
> are addressed. If you have received this email in error please notify
> the system manager.
>
> This footnote confirms that this email message has been swept
> for the presence of computer viruses and is believed to be clean.
>
> **********************************************************************
>
> ------------------------ MailScanner list ------------------------
> To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
> 'leave mailscanner' in the body of the email.
> Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
> the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
>

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).