bayes expire tokens

[David Lee]

On Tue, 8 Mar 2005, Julian Field wrote:

> Looks like I need to take another look at this again.
> One process doing a Bayes rebuild should lock out all the other
> processes from trying to use SA, or it should make them wait.
> What is the value of your "Wait For Bayes Rebuild" setting?
>
> It may be a little while until I manage to fix this, but it is top of
> the list of problems to be resolved.

Thanks, Julian.  That would be appreciated by many people.

A bit more detail, in the hope that it helps understand the problem.

This problem used to plague us.  I massively reduced, but not eliminated,
it a few months ago.

I know that various folk on the list suggest switching off the MS-driving
of this procedure (change "Rebuild Bayes Every" from 3600 to 0 and instead
use a cron job).  But this always felt suboptimal (to me, in our context).

Our settings had been:
    SpamAssassin Timeout = 40
    Wait During Bayes Rebuild = yes

and we got bucketloads of these part-built "expire" files every day.  At
the suggestion of various folk, I increased "SpamAssassin Timeout" to 120
(two minutes), and the number of these orphaned files dramatically fell
(although not quite to zero).

As an interim workaround, this latter (increase "SpamAssassin Timeout"),
though imperfect, felt better (to me, in our context) than the former
(external, bolt-on agency of "cron") in that it was using MailScanner as
intended.

But a proper fix would be wonderful!  (I'd offer to dabble in the code
myself, but I don't know it well enough.  But I'd be happy to test any
proposed fixes you might have.)

Hope that helps.


--

:  David Lee                                I.T. Service          :
:  Senior Systems Programmer                Computer Centre       :
:                                           University of Durham  :
:  http://www.dur.ac.uk/t.d.lee/            South Road            :
:                                           Durham                :
:  Phone: +44 191 334 2752                  U.K.                  :

------------------------ 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).

Support MailScanner development - buy the book off the website!