On Tuesday 13 January 2004 01:45 pm, Lucas Albers wrote: > Brett Simpson said: > > I think I figured it out. I needed to increase S:45s;R:45s to a higher > > value > > since it was 45 seconds between the start of the connection and the > > error. > > > > INPUT_MAIL_FILTER(`mimedefang', > > `S=unix:/var/spool/MIMEDefang/mimedefang.sock, > > F=T, T=S:5m;R:5m;E:15m') > > > Your slaves are timing out, correct, and upping the timeout fixed it? > Do you have some strange filter that are using up the time?
I was getting errors a couple times an hour. After making the change I have received no timeout errors even with the network (RBL, Razor, DCC, Pyzor) checks enabled. My SpamAssassin Bayes files are 7MB in size and I'm using several large .cf files (like BigEvil.cf). So I guess it's SpamAssassin that's taking to long to respond. Also placing the Bayes journal on a ram disk and disabling the network checks didn't help. Brett _______________________________________________ Visit http://www.mimedefang.org and http://www.canit.ca MIMEDefang mailing list [EMAIL PROTECTED] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
