RE: [Mimedefang] Getting Error from multiplexor: ERR No response from slave
Do all these rules have pretty much the same value? Can they be combined into single rules that match on multiple strings? Just quadruple your timeouts and see if it can handle all those rules. Lucas, The script produces descrete rawbody rules for each line of text in the input file. As for point-value, I am originally setting everything to a point value of zero. My intent was to observe the performance impact of such a huge set of rules, without (yet) letting the rules otherwise influence the message in any way. Once we knew how expensive the rules were, the intent was to score them high enough to reject the message (since that's what they do on the Exchange boxes internal to my organization). If the input file looks like this: line one line two line three The resulting rules look like this: rawbody LOCAL_EMGR_STRING_1 /line one/i describe LOCAL_EMGR_STRING_1 Unacceptable word or phrase scoreLOCAL_EMGR_STRING_1 0 rawbody LOCAL_EMGR_STRING_2 /line two/i describe LOCAL_EMGR_STRING_2 Unacceptable word or phrase scoreLOCAL_EMGR_STRING_2 0 rawbody LOCAL_EMGR_STRING_3 /line 4/i describe LOCAL_EMGR_STRING_3 Unacceptable word or phrase scoreLOCAL_EMGR_STRING_3 0 Each line of text in the input file represents a string or phrase that on the Exchange servers, is considered evil enough to cause message rejections. There is no pattern or other relationship between the lines of input... they're just a list of dirty words, phrases, and other no-no patterns. KEN CORMACK, RHCE Sr. UNIX Systems Analyst, Open Systems Group Sr. Software Analyst, TSG Midrange Systems Group AFFILIATED COMPUTER SERVICES, INC. ___ Visit http://www.mimedefang.org and http://www.canit.ca MIMEDefang mailing list [EMAIL PROTECTED] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
RE: [Mimedefang] Getting Error from multiplexor: ERR No response from slave
At 05:59 AM 3/29/2004, Cormack, Ken wrote: As for point-value, I am originally setting everything to a point value of zero. My intent was to observe the performance impact of such a huge set of rules, without (yet) letting the rules otherwise influence the message in any way. As I understand it, setting the score of a SpamAssassin rule to 0 disables the rule. For testing, small values like 0.01 are recommended. If all these scores are set to 0, SA isn't using them, so something else may be going on. If you haven't already, you should run spamassassin -D --lint and mimedefang.pl -test to make sure there are no syntax errors lying in wait. Kelson Vibber SpeedGate Communications www.speed.net ___ Visit http://www.mimedefang.org and http://www.canit.ca MIMEDefang mailing list [EMAIL PROTECTED] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
RE: [Mimedefang] Getting Error from multiplexor: ERR No response from slave
Kelson, You've solved the problem... As I understand it, setting the score of a SpamAssassin rule to 0 disables the rule. For testing, small values like 0.01 are recommended. Excellent point. I've raised the value to 0.01 for my tests. If you haven't already, you should run spamassassin -D --lint and mimedefang.pl -test to make sure there are no syntax errors lying in wait. Running spamassasin -D --lint caught a few occurrences of special characters in the stings, that seemed to be tripping up the SA engine's regex parser. Until I get those characters properly escaped, I removed those with grep -v and re-ran the -D --lint with those entries removed. The test results were clean. top shows the size of my multiplexor threads has grown from approx 20M each, to now about 75MB each. I'm running it with embedded perl mode, with the lion's share of that footprint in shared memory, so it's not as bad as it sounds. :) Consequently, with the rules now free of regex problems, processing appears to be doing fine, and I am no longer seeing the timeout errors I was getting before. Thanks for the suggestion. KEN CORMACK, RHCE Sr. UNIX Systems Analyst, Open Systems Group Sr. Software Analyst, TSG Midrange Systems Group AFFILIATED COMPUTER SERVICES, INC. 557 E. Tallmadge Ave., Akron, OH 44310 mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] Phone: (330) 643-6372 Fax: (330) 643-6367 Pager: (800) 946-4646 Pin 1412819 If that that is 'is' is that that is not 'not is', is that that is 'not is' that that is not 'is'? It is! - Ken Cormack Sendmail administration is not black magic. There are legitimate technical reasons why it requires the sacrificing of a live chicken. - Unknown ___ Visit http://www.mimedefang.org and http://www.canit.ca MIMEDefang mailing list [EMAIL PROTECTED] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
