4.26.7, bayes rebuild, confused. -- 4.26.8
Jeff A. Earickson
jaearick at COLBY.EDU
Mon Feb 2 15:08:29 GMT 2004
Popped 4.26.8 into place, let it run for a few minutes (Sol 9, SA
2.63, Razor). No more complaints about Bayes, but no SpamAssassin
messages either. I ran a batch in debug mode for both MS and SA, and
it looked like stuff in the debug batch got tagged by SA:
debug: is spam? score=10.95 required=5 tests=BAYES_20,FROM_ENDS_IN_NUMS,MIME_MISSING_BOUNDARY,MISSING_MIMEOLE,MSGID_FROM_MTA_SHORT,NO_REAL_NAME,PRIORITY_NO_NAME,RAZOR2_CF_RANGE_51_100,RAZOR2_CHECK
but nothing in the syslog regarding SA. I also set the log level
for razor to 4 and razor is busy. How to check it 4.26.8 is really
using SA, if nothing appears in syslog? I'm back to running 4.25-14.
On Mon, 2 Feb 2004, Julian Field wrote:
> Date: Mon, 2 Feb 2004 14:06:40 +0000
> From: Julian Field <mailscanner at ECS.SOTON.AC.UK>
> Reply-To: MailScanner mailing list <MAILSCANNER at JISCMAIL.AC.UK>
> To: MAILSCANNER at JISCMAIL.AC.UK
> Subject: Re: [MAILSCANNER] 4.26.7, bayes rebuild, confused. -- 4.26.8
> I have just posted version 4.26.8.
> The problem did not appear on Linux, but does appear on Solaris. You can
> now disable all the relevant code by setting
> Rebuild Bayes Every = 0
> I will look into fixing this as a priority, but it is highly OS-specific
> and may even be Perl-version specific. It refuses to lock a file it has
> just successfully opened, but seems happy when I do it elsewhere :-(
> P.S. thanks for your patience....
> At 10:23 02/02/2004, you wrote:
> >Inevitably I put a config name wrong in that one.
> >At 09:40 02/02/2004, you wrote:
> >>Please could you try the attached SA.pm and see if it helps.
> >> - Set "Rebuild Bayes Every = 0" should disable all this code.
> >> - Locking code changed to more closely match the virus scanner
> >>locking code.
> >>The trouble is, it all works for me. But that's on a Linux system, and the
> >>underlying locking behaviour may well be different on Solaris.
> >>At 07:56 02/02/2004, you wrote:
> >>>I have the same behaviour with the rebuild of bayes database and I get it
> >>>every time MailScanner is launched.
> >>>To avoid the "Skipping", I have to "manually" remove the lock file
> >>>(for me it's not important since I do not use bayes !)
> >>>In SA.pm, the lock file is created before the test on "$RebuildBayes"
> >>>and the lock is removed only if the bayes database has been rebuild.
> >>>If $RebuildBayes == 0, the lock will never be removed.
> >>>if $RebuildBAYES <> 0, it seems that the rebuild on solaris don't finish
> >>>or don't begin ???
> >>> I see the "Skipping" line in the logfile but I don't see any line
> >>> such as "SpamAssassin Bayes database rebuild preparing" even with
> >>> $RebuildBAYES <> 0
> >>>-- Pascal --
> >>> --
> >>Julian Field
> >>MailScanner thanks transtec Computers for their support
> >>PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654
> >Julian Field
> >MailScanner thanks transtec Computers for their support
> >PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654
> Julian Field
> MailScanner thanks transtec Computers for their support
> PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654
More information about the MailScanner