>On Fri, 13 Jan 2006 12:40:14 +0100, you wrote: >George R. Kasica schrieb: > >> Help needed please! >> >> We are trying to upgrade from exim 3.36 and SA 3.0.4 to Exim 4.60 and >> SA 3.1 and are having no end of difficulties here. We just decided to >> pull SA 3.1 out and go back to 3.0.4 as we cannot get it to scan each >> message, not time out or crash and not use up all the CPU cycles. >> without SA running load with a w is generally < 2 with it up and going >> 10, 12 or higher is not unusual and causes many problems. 3.0.4 does >> not have this problem. We have cut our rules files down from the SARE >> set we are running to the "stock" set from the 3.1 install with little >> difference - it still times out and skips mail scans here for no >> obvious reason. The latest glitch is that we are sending out multiple >> copies of e-mails I'm thinking due to system load caused by SA 3.1 >> messing up Exim. > >Well, without relevant configuration (SA invocation) and log snippets we >won't be able to tell you what happens on your machine. But in the past, >people who reported similar problems piped every message into SA. Since >SA takes quite some time processing a message, specially bigger ones, >exim loses patience, assumes SA is down and returns a temporary error to >the sender. To avoid this, you should limit the size of messages that >you want to be content-scanned with SA. There's no need to check whether >a 1 MB message is spam. We are only feeding messages smaller than 250KB >to SA and haven't had any problem with SA timeouts yet. > >HTH, >Patrick Eisenacher
Pat: Things are looking back to normal here with 4.60 and 3.1 now after some help by Larry Rosenman <ler@lerctr.org>, turns out that one of the SARE rules (I'm thinking one of the blacklist URI ones about 13 meg in size (sorry I don't recall the name of it at the moment) may have been causing our grief along with not being on Perl 5.8.7 which we also did here. After cleaning up those 2 items it is looking good so far. Now to slowly add back the rest of the SARE stuff and reactivate Bayes and see how we do. Thanks for the response, George ===[George R. Kasica]=== +1 262 677 0766 President +1 206 374 6482 FAX Netwrx Consulting Inc. Jackson, WI USA http://www.netwrx1.com [EMAIL PROTECTED] ICQ #12862186
