4.30.3 sa counting error

[Julian Field]

At 20:15 05/05/2004, you wrote:
>Julian,
>    With your patch in place, what I see now is:
>
>SpamAssassin timed out and was killed, failure 0 of 20
>
>no number higher than zero there, shouldn't it start at one?
>Only two instances today so far...

That is theoretically possible. What I do now is run a queue which stores
the last n success/failure codes. It keeps track of the total of the
members of the queue. If one got put in at the same time as the last one
went out, I think it could hit zero on a failure. Not 100% sure though, I
admit :-)

The aim was to detect RBLs and SA failing, say, 50% of the time, by setting
a total timeout limit and the length of the queue. So if 10 out of 20 tests
succeeded, it would know it failed 50% of the time which might be enough to
kill it off if the thresholds were set to 10 out of 20. You can get RBLs,
in particular, that start failing most of the time, but not all the time.
So the old test for n consecutive failures wouldn't trigger, but your mail
throughput slowed to a crawl.
--
Julian Field
www.MailScanner.info
Professional Support Services at www.MailScanner.biz
MailScanner thanks transtec Computers for their support
PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654

-------------------------- MailScanner list ----------------------
To leave, send    leave mailscanner    to jiscmail at jiscmail.ac.uk
Before posting, please see the Most Asked Questions at
http://www.mailscanner.biz/maq/     and the archives at
http://www.jiscmail.ac.uk/lists/mailscanner.html