Re: V3.1.2 hanging up

2006-05-29 Thread Loren Wilton
> 2) Occasionally, a child process will spawn off and sit there using 60-80% > of the CPU for 20-30 (or more) minutes. Each time I've seen it thus far, > I've ended up killing the process. Are you running bayes or awl? If so, check the database sizes. It might be the child is trying to do a hu

Re: .spamassassin

2006-05-29 Thread George Georgalis
On Mon, May 29, 2006 at 09:13:15PM -0400, Daryl C. W. O'Shea wrote: > >So you're using spamd. From perldoc spamd: > >-x, --nouser-configDisable user config files > OK. that works. I note that I must also set: use_bayes 0 #use_auto_whitelist 0 disabling use_auto_whitelis

Re: .spamassassin

2006-05-29 Thread Daryl C. W. O'Shea
On 5/29/2006 9:06 PM, George Georgalis wrote: On Mon, May 29, 2006 at 07:20:31PM -0400, Daryl C. W. O'Shea wrote: On 5/29/2006 5:20 PM, George Georgalis wrote: Looking at this block from above SpamAssassin.pm line 1469, it's not clear to me how to avoid the warning/errors listed below. note:

Re: .spamassassin

2006-05-29 Thread George Georgalis
On Mon, May 29, 2006 at 07:20:31PM -0400, Daryl C. W. O'Shea wrote: >On 5/29/2006 5:20 PM, George Georgalis wrote: >>Looking at this block from above SpamAssassin.pm line 1469, >>it's not clear to me how to avoid the warning/errors listed below. >> >>note: I'm doing active (in smtp) filtering, and

NONELEMENT Scoring

2006-05-29 Thread Dan
I'm trying to understand the scoring for HTML_NONELEMENT (20_html_test.cf).  There seems to be a sweet spot between 70 and 80%, but its not clear why 80-100% isn't scored:score HTML_NONELEMENT_00_10 0score HTML_NONELEMENT_10_20 0score HTML_NONELEMENT_20_30 0score HTML_NONELEMENT_30_40 0score HTML_N

Re: .spamassassin

2006-05-29 Thread Daryl C. W. O'Shea
On 5/29/2006 5:20 PM, George Georgalis wrote: Looking at this block from above SpamAssassin.pm line 1469, it's not clear to me how to avoid the warning/errors listed below. note: I'm doing active (in smtp) filtering, and I do not want to create a .spamassassin directory. You don't say how, but

Re: bypassing SPF check for some domains

2006-05-29 Thread Daryl C. W. O'Shea
On 5/26/2006 8:58 PM, klaus thorn wrote: Hi spamassassin crowd, can I configure spamassassin to not do SPF check for certain domains? No, but if you don't want to check mail against your own SPF records you could use split views in your DNS... with your internal (or mail server facing) view

Re: quick 3.1.2 questions

2006-05-29 Thread Rick Macdougall
[EMAIL PROTECTED] wrote: I just upgraded from 3.0.4 and everything went ok, except: LINT complains about one thing it never did before in local.cf: bayes_path /var/spamassassin/ Hi, bayes_path /var/spamassassin/bayes Regards, Rick

.spamassassin

2006-05-29 Thread George Georgalis
Looking at this block from above SpamAssassin.pm line 1469, it's not clear to me how to avoid the warning/errors listed below. note: I'm doing active (in smtp) filtering, and I do not want to create a .spamassassin directory. Ideas? // George sub get_and_create_userstate_dir { my ($self) =

quick 3.1.2 questions

2006-05-29 Thread up
I just upgraded from 3.0.4 and everything went ok, except: LINT complains about one thing it never did before in local.cf: bayes_path /var/spamassassin/ produces: [15987] warn: config: SpamAssassin failed to parse line, "/var/spamassassin/" is not valid for "bayes_path", skipping: bayes_p

RE: Razor2 Error

2006-05-29 Thread Dan O'Brien
"Jim Hermann - UUN Hostmaster" <[EMAIL PROTECTED]> wrote: > I found my problem. I had some custom definitions that used > eval:check_razor2_range. I deleted the custom definitions > and the error went away. > > Jim Yup... that was my problem, too: # Razor Adjustments score RAZOR2_CF_RANGE_5

[FOLLOW-UP] End-user (not administrator) question

2006-05-29 Thread Ricardo P. Jasinski
Hi fellows, thanks to everyone who bothered sharing their opinion on this matter. Loaded with your suggestions, I called our administrator and kindly requested to be added to the spam-lovers list. He just asked my motivations, wich I kindly explained, and he quickly made the change. Note: before

Re: Penalizing mail with multiple Subject or From header fields?

2006-05-29 Thread Mark Martinec
> header L_DOUBLE_SUBJECT ALL =~ /^Subject:.*^Subject:/smi > header L_DOUBLE_FROMALL =~ /^From:.*^From:/smi Doing some more research on my suggestion and evaluating results for the last three days, here is a proposal for new rule: http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4927

RE: Razor2 Error

2006-05-29 Thread Jim Hermann - UUN Hostmaster
I tried changing the definitions to the new format. I still got the error. # complete set of ranges - jwh 9/29/05 bodyRAZOR2_CF_RANGE_00_01 eval:check_razor2_range('','00','01') tflags RAZOR2_CF_RANGE_00_01 net describe RAZOR2_CF_RANGE_00_01 Razor2 gives confidence between 00 and 01 bo

V3.1.2 hanging up

2006-05-29 Thread Will Nordmeyer
Hi, I installed V3.1.2 last night on my Sendmail/Procmail based system. Everything (mostly) seems to be running fine, but I've seen a few troublesome issues... 1) At 3am (or so) last night - spamd just died. The system went for 4 hours without any spamchecking. 2) Occasionally, a child proces

Re: bypassing SPF check for some domains

2006-05-29 Thread Magnus Holmgren
On Saturday 27 May 2006 02:58, klaus thorn took the opportunity to write: > [EMAIL PROTECTED] is sending to > [EMAIL PROTECTED] Andy is somewhere (home/office/congress) > using often changing IP addresses. To compansate for this, > he uses the server example.com to relay all mail, > (authenticating

bypassing SPF check for some domains

2006-05-29 Thread klaus thorn
Hi spamassassin crowd, can I configure spamassassin to not do SPF check for certain domains? I assume that spf whitelisting is doing something else: after a successfull spf check, classify the mail as ham. Why I want to do this: [EMAIL PROTECTED] is sending to [EMAIL PROTECTED] Andy is somewher

Re: Razor2 Error

2006-05-29 Thread Daryl C. W. O'Shea
Jim Hermann - UUN Hostmaster wrote: I found my problem. I had some custom definitions that used eval:check_razor2_range. I deleted the custom definitions and the error went away. What were the custom rules you had that caused that error? It's likely that we could add/improve validation of t