bayes expire tokens

[Chris Conn]

    [ 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. ]

Kai Schaetzl wrote:
> David Lee wrote on         Wed, 9 Mar 2005 10:04:39 +0000:
>
>
>>Thanks, Julian.  That would be appreciated by many people.
>>
>
>
> I don't see that Julian can do much or even little about this. The problem
> happens because the bayes_expiry takes too long. This can only happen on a
> corrupted db or a large db. If you get quite a few messages (I think it
> should be at least a few 10.000 per day, I don't have such a system, I
> can't compare) running thru your system the latter is likely the case.
> There are two options then: stop bayes_expiry at all (which will keep it
> growing, but may work out) or do it on a scheduled basis. You can also
> tweak the db size either to be very small (reducing the amount of time it
> takes to expire) or rather large (so that the automatic expiry is unlikely
> to happen). But this is all SA-related, not MS. Please go over to the SA
> list and talk about your problems. If people just keep whining here,
> nothing will change, since the developers don't know that it happens so
> frequently. And if the problem is not related to what I think the SA list
> is better suited for that anyway.

Hello and thank you for your helpful perspective...Have you read what Mr
Field has replied to me?  I believe you are missing the point:

Here is the quote:

"
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?
"

The logs I sent show that another MailScanner child process is still
analyzing mail during a rebuild.

Now I don't see anyone whining here, other than your own flame.  I have
a cron job that deletes expire files older than 2 days,  so I don't
personally care if this is resolved in a distant future.  We are
reporting factual events.

And I disagree that this is SA-related as you claim, since the process
actually calling expiry is MailScanner.

Don't get me wrong, I think MailScanner does an excellent job.  But what
I and some others are pointing out, is that bayes use occurs even when
MailScanner is expiring; even if it takes an hour to expire, the
configuration of MailScanner (not SA) suggests you can pause MailScanner
while this happens.  To me and to others, including Mr Field, this is
not exactly what happens.

Chris

------------------------ 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!