On Fri, 29 Jul 2016 21:13:56 +0200 Robert Schetterer <r...@sys4.de> wrote:
> so i.e measure mails tagged as spam by spamassassin > with pure greylisting setup running before tagging ,perhaps for one > week, then stop greylisting ,do the same with pure postscreen setup, > compare results, this way you may given direction if you still need > greylisting. See the attached reports. The first one shows that the vast majority of greylisted messages are not retried. The second shows that greylisting stops a pretty high percentage of messages compared to spam detection (the "Spam" and "Quarantined" categories) and is therefore an effective technique that can greatly reduce the processing burden. The Y-axis indicates messages per hour. Regards, Dianne.
