On Wed, 2007-10-24 at 20:31 +0200, Christian Rehkopf wrote: > nice idea to show stats - its really interesting to see how it works for > others.
Yeah I figured more info might help rather than less. Not that stats are complete info, but helps. > As I didnt like the subject of this therad here I changed it a bit :) Yeah that might have come across as a little to negative. Didn't mean to be :) > You recieved ~ 20k Connections - but only 582 messages passed in the end > - great! > > just the numbers inbetween differs from mine alot (in percentage) - so > it needs just to be tweaked at the settings. Ok, haven't had to do that in the past. Mostly because it was so effective with the defaults. I didn't want to mess with that by trying to squeeze every last drop from the prune, and make it less effective instead of more. > I dont use bayens, because I have a higher load of messages and it made > the CPU Load to high, That's not so much problem/gripe. For me it's the amount of bayesian since I run daily reports on it. To make sure no valid ones are rejected, since that occurs most times by bayesian. They used to be big, and long ago when I upgraded to 1.2.x they were cut at min in half. Which made me very happy :) I just expected a bit more from 1.3.x based on the following statements. http://sourceforge.net/mailarchive/message.php?msg_name=fc.000f455505ff6da23b9aca00fe4e1873.5ff6db2%40mail.vnet.de It could be that I am already in that range only 5%-10% are hitting the bayesian filter now. But that didn't change from 1.2.x -> 1.3.x. > but prefer RBL & RWL. The point is that u neet a > fast DNS for that. I run two local DNS servers. It wasn't the DNS aspect that took a while, it was contacting the third party RBLs or etc that caused a delay. At times close to like 1 minute or more. I didn't find it acceptable considering it didn't seem to make a diff. Bayesian spam did not drop, nor did any other stats. > At least I run alost all Header based tests active - and ist running > excellent and without a notifiable CPU-Load. Not sure what I have there. > That "denied SMTP conns" is zero in your stats may be caused in not > using the setting "denySMTP" or not rebuilding spam DB, or not using the > PB. Rebuild spam db occurs nightly via cron job for years now. It's part of my daily monitoring of assp, and my asspr reports are called and generated from same script that rebuilds. PB is enable and working. I can see it in the logs as I tail -f the log file. I wanted PB and SMTP limiting long ago, and was stoked to see those features added. Those plus greylisting were EXTREMELY effective from day one. Those three features along, cut my bayesian stuff in half or less than. > The features ASSP is offering for catching SPAM already in headers are > unbelieveable, but need to be understanding of how its running. That I doubt I have any understanding of. But now that it's mention. I will surely look into it. Really appreciate your input and feedback. > > jfyi > http://www.crnet.de/labor/assp/index.htm Thanks reviewing and diffing to mine. -- William L. Thomson Jr. Gentoo/Java
signature.asc
Description: This is a digitally signed message part
------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
