Sorry, yeah - all BAYES rules were gone after --clear. When I trained it with another fresh 1000 spam and ham it started again with BAYES_00. I will need to go through the spam and ham again
On 11 September 2015 at 23:26, RW <rwmailli...@googlemail.com> wrote: > On Fri, 11 Sep 2015 22:25:46 +0100 > Peter Kelly wrote: > > > > I can actually see the 0.0 scores directly in the logs > > > I tested one and it was out, but only by 0.04. I thought it was > probably due to your cron job running a bit too early for this morning's > update. > > > > It must be the bayes is completely messed up. I actually did a > > sa-learn --clear and ran a fresh 1000 spam and ham (verified) through > > again, still seeing BAYES_00 an unusually high number of times. > > Are you sure you ran that on the right database? With an empty > database you should see log entries without any BAYES_* rules. > > > I will look into my own DNS nameserver for the RELAY and TRUSTED > > problem. > > ALL_TRUSTED doesn't have anything to do with DNS, it's determined from > the received headers. spamassassin -D is useful for debugging it. >