lst_ho...@kwsoft.de put forth on 12/7/2010 2:20 PM: > Zitat von Stan Hoeppner <s...@hardwarefreak.com>: > >> lst_ho...@kwsoft.de put forth on 12/7/2010 2:18 AM: >>> Zitat von Stan Hoeppner <s...@hardwarefreak.com>:
>> Postgrey's auto whitelist feature is on by default. In fact, you can't >> disable it. If you could, Postgrey would break, as it would never find >> a triplet. > > Huh?? > > --auto-whitelist-clients=N whitelist host after first successful delivery > N is the minimal count of mails before a > client is > whitelisted (turned on by default with > value 5) > specify N=0 to disable. Sorry, got my terminology screwed up--we were talking about two different things. > We use --auto-whitelist-clients=1 with long initial delay and very long > max-age. With this nearly all of the mailserver we every speak with are > whitelisted. The other ones are (legal) advertisment which can wait anyway. > >> I'm not going to delve into the reasons why greylisting sucks. We all >> already know them. Some of us can tolerate them, some can't. Some >> could for a while (me) and got fed up given the very low RIO here. >> >> Everyone's mileage varies. If you're having good results with Postgrey >> you should try the fqrdns.pcre file that I recommended in the thread you >> replied to. You'll likely see very quickly why greylisting is redundant >> while using that file. > > Sorry, but we have customer operating on lines with very strange DNS > settings if it even could be called so. I don't have the time to explain > (r)DNS to some clueless remote. So in our case Greylisting is the > minimal hassle solution cutting some ten thousands spams a day with zero > FP. I can see that point of view. However, the file I mentioned only targets specific rDNS patterns of specific ISPs, residential mainly, and some small business where the ISP overlaps the two (but shouldn't). -- Stan
