Andy Figueroa wrote:
> I'm growing increasingly wary of running with AWL enabled (I am).
> When I see AWL invoked on particular messages, the score seems to push
> as often in the wrong direction as in the nominal right direction.
>
> What is the evolving conventional wisdom regarding using AWL?
(cc'ing users list since there's nothing private in here)
On Fri, Jan 19, 2007 at 09:55:14PM -0500, Brian Z wrote:
> I thought it was safe to run spamd as root?
Running the spamd parent as root is fine, but the children won't run
a message through as root as a security paranoia measure. So if ro
On Fri, Jan 19, 2007 at 03:59:29PM -0800, Kenneth Porter wrote:
> I just got a message that had
> image/jpeg as its top-level MIME component.
0.068 0.0776 0.1.000 0.500.00 TVD_WHAA_CONTENT
This looks for mails that have a Content-Type specified, and it's not text or
multipar
On Fri, Jan 19, 2007 at 06:01:06PM -0500, tekmonster wrote:
> After restarting though I now get the following error
These have nothing to do with sa-update fwiw.
> Jan 19 17:55:05 nj1hqls01 spamd[19756]: spamd: connection from localhost
> [127.0.0.1] at port 50439
> Jan 19 17:55:05 nj1hqls01 sp
On Fri, Jan 19, 2007 at 02:50:10PM -0800, nigel wrote:
> so if it created a DB for me, how do i get rid of the 4.4GB of tokens that
> exist in /var/spool/MailScanner/spamassassin?
You need to properly use the right DB path. Either a) run as the correct
user, b) if you have privs use --dbpath, c)
Theo Van Dinter wrote:
(I assume this message was also supposed to goto the users list since there
was nothing private in it, so cc'ing there.)
On Fri, Jan 19, 2007 at 03:33:27PM -0500, [EMAIL PROTECTED] wrote:
The thought I was struggling with is that in the MTA the content_filter
is told who
I am running fedora 6 and it spamd seemed to be running ok.
After restarting though I now get the following error
Jan 19 17:55:05 nj1hqls01 spamd[19756]: spamd: connection from localhost
[127.0.0.1] at port 50439
Jan 19 17:55:05 nj1hqls01 spamd[19756]: spamd: setuid to root succeeded
Jan 19 17
Are there any rules in the current release (or in updates) to score a
message that contains no text parts? I just got a message that had
image/jpeg as its top-level MIME component. But it's almost as bad to get a
multipart that contains neither text/plain nor text/html (or any other text
varian
so if it created a DB for me, how do i get rid of the 4.4GB of tokens that
exist in /var/spool/MailScanner/spamassassin?
thanks
Nigel
Theo Van Dinter-2 wrote:
>
> On Fri, Jan 19, 2007 at 12:04:19PM -0800, nigel wrote:
>> debug: failed to load Net::DNS::Resolver: Can't locate Net/DNS.pm in @IN
I'm growing increasingly wary of running with AWL enabled (I am). When
I see AWL invoked on particular messages, the score seems to push as
often in the wrong direction as in the nominal right direction.
What is the evolving conventional wisdom regarding using AWL?
Andy Figueroa
(I assume this message was also supposed to goto the users list since there
was nothing private in it, so cc'ing there.)
On Fri, Jan 19, 2007 at 03:33:27PM -0500, [EMAIL PROTECTED] wrote:
> The thought I was struggling with is that in the MTA the content_filter
> is told who the email is going to
On Fri, Jan 19, 2007 at 03:09:06PM -0500, [EMAIL PROTECTED] wrote:
> I was looking at using spampd for my content_filter in postfix and found
> a comment in the debian package description that this is not suitable
> for use with a per-user Bayesian filtering configuration.
>
> I wanted to confirm
I was looking at using spampd for my content_filter in postfix and found
a comment in the debian package description that this is not suitable
for use with a per-user Bayesian filtering configuration.
I wanted to confirm that this was still valid or not.
I eventually intend on setting up a datab
On Fri, Jan 19, 2007 at 12:04:19PM -0800, nigel wrote:
> debug: failed to load Net::DNS::Resolver: Can't locate Net/DNS.pm in @INC
You should get Net::DNS properly installed. It's good stuff.
> debug: config: SpamAssassin failed to parse line, skipping: ARE "Adult"
> Ruleset for SpamAssassin
Mi
here is the debug output
[EMAIL PROTECTED] postfix]$ sa-learn -D --force-expire
debug: SpamAssassin version 3.0.1
debug: Score set 0 chosen.
debug: running in taint mode? yes
debug: Running in taint mode, removing unsafe env vars, and resetting PATH
debug: PATH included '/usr/kerberos/bin', keepi
I'll give that a try.
I know that this might sound trivial to some but we recently added a
rule to prod and our message processing time when from 1.2 seconds to
2.6 seconds, which doesn't sound like a lot but during peak we are
processing about 60k messages an hour (across 4 frontend machines).
G
On Fri, Jan 19, 2007 at 11:31:49AM -0800, Gary W. Smith wrote:
> With that said, does anyone know how I would run 5 different copies of
> spamd on the same box?
Run on different ports w/ different startup parameters pointing at different
config copies.
> Better question, what's the best way to
> > More specifically, I would like to test different threshold levels,
> > different rule sets, tie some of the instances into lookup's, etc.
My
> > goal is to process each incoming email against each of these
instances
> > to see what the primary differences are.
>
> It sounds more like you wa
On Fri, Jan 19, 2007 at 11:17:37AM -0800, Gary W. Smith wrote:
> For the purposes of development, I would like to run 5+ separate copies
> of spamd on the same host to test different settings. I would like to
> be able to have a different config for each instance. It this possible?
Sure.
> More
On Fri, Jan 19, 2007 at 10:07:53AM +, Justin Mason wrote:
> > Thank you. That explains it then. I suppose there is no fix for the false
> > positives then? Since its the total message size. I began thinking that
>
> 1.6 points is not an FP. it's expected that some rules *will* fire
> on
For the purposes of development, I would like to run 5+ separate copies
of spamd on the same host to test different settings. I would like to
be able to have a different config for each instance. It this possible?
More specifically, I would like to test different threshold levels,
different rul
On Fri, Jan 19, 2007 at 09:17:31AM -0800, nigel wrote:
> i've just ran this and it did not delete any of the tokens. i have saved the
> output (i used it in debug mode) if that will help
Feel free to pastebot the debug output. FWIW, --force-expire doesn't
guarantee any tokens will be expired, it
i've just ran this and it did not delete any of the tokens. i have saved the
output (i used it in debug mode) if that will help
Nigel
nigel wrote:
>
> ok
>
> how long would you expect this to take? its running on an old server
> which doesnt help :-)
>
> Nigel
>
>
>
> Martin.Hepworth w
I'm not sure if this is a MailScanner problem or a SpamAssassin problem,
but someone here will at least be able to help me narrow it down. I am
running MailScanner 4.57.6 on Suse Linux 10.1. My MTA is Postfix 2.3.6
and I'm running SpamAssassin 3.1.7.
The problem I seem to be having is that
ok, first do i have to stop MailScanner? Secondly, if i make that change i
guess i would have to run "sa-learn --force-expire" to make it clean up
there and then? if i didn't run that command it would do it after the time
period set.
am i correct here? :)
Nigel
Martin Hepworth-3 wrote:
>
>
Nigel
Then in your case you've not told anything to clean up the bayes
files...so yes try the inbuilt MailScanner way
First time it may take some time as you've got so many files, so you
make wish to stop MailScanner, run "sa-learn --force-expire" (as the
appropriate user if non-root) and then re
ok
how long would you expect this to take? its running on an old server which
doesnt help :-)
Nigel
Martin.Hepworth wrote:
>
> Nigel
>
> Yes, but given you've got lots of expiring to do I suggested you run
> the sa-learn by hand first, then you have some feedback for whats
> happening, rat
Nigel
Yes, but given you've got lots of expiring to do I suggested you run
the sa-learn by hand first, then you have some feedback for whats
happening, rather than MailScanner just waiting..
--
Martin Hepworth
Snr Systems Administrator
Solid State Logic
Tel: +44 (0)1865 842300
> -Original Me
my mailscanner settings are slightly different
Rebuild Bayes Every = 0
Wait During Bayes Rebuild = no
so if i change them to what you suggested this will clear them out? also,
does this effect mail delivery while it is doing it?
Nigel
Martin Hepworth-3 wrote:
>
> Nigel
>
> List is best..
>
Office of the Postmaster writes:
>
> > > A 3-5% ratio of our email is getting tagged with HTML_IMAGE_ONLY_32 which
> > > according to the DOC is HTML: images with 2800-3200 bytes of words. The
>
> >Yes, it's the message size that's in question, not the "" tag. it's
> >expected that the rule may
Nigel
List is best..
Commented out is fine - as in that's what I've got.. ;-)
If you're other settings in MailScanner.conf are also as below then
you'll have to use the cron method..
--
Martin Hepworth
Snr Systems Administrator
Solid State Logic
Tel: +44 (0)1865 842300
> -Original Message-
Try
spamassassin --lint
(two dashes, not one).
Cheers,
Phil
--
Phil Randal
Network Engineer
Herefordshire Council
Hereford, UK
_
From: D Ivago [mailto:[EMAIL PROTECTED]
Sent: 19 January 2007 09:26
To: users@spamassassin.apache.org
Subject: Re: blacklisting
*EVERY* time
Nigel
Ideally this would be on the mailscanner list...
There's a couple of ways MailScanner can expire the bayes tokens down,
and which one works for you seems to be hit and miss from my
experience..
1st method is to use the 'internal' MailScanner method. In
MailScanner.conf, make sure
Rebuild
* Rocco Scappatura <[EMAIL PROTECTED]>:
>
> Hello,
>
> I use amavisd-new. When I send emails from Lotus Notes they get blocked.
Why? What's in the logs?
--
Ralf Hildebrandt (i.A. des IT-Zentrums) [EMAIL PROTECTED]
Charite - Universitätsmedizin BerlinTel. +49 (0)30-450 570
Hello,
I use amavisd-new. When I send emails from Lotus Notes they get blocked.
Even If they are "plain" messages. Indeed they are however MIME
messages.
I would like to verify if there is a way to analyse what is the tokens
whose raise the score so that the message is considered spam while the
hi everyone, i'm pretty new to SpamAssassin. Disk space on one of our
servers is running low and it seems that /var/spool/MailScanner/spamassassin
is using 4.4GB. it is full of items like this
bayes_toks.expire7000
i roughly know what they do but is there a way to clear this out? i thought
so
*EVERY* time you edit your config files, run spamassassin --lint. It
should run quietly and just exit if all is well. If there's a problem
parsing your config, it will print a message to that effect.
so I added
blacklist_from [EMAIL PROTECTED]
to my local.cf and ran 'spamassassin -lint' but
Hello, all.
I would like to set different SPF rule per domain.
For example,
-. I would like to check SPF only per "example.com" or "abc.net".
-. I would like to set different SPF score for "abc.com" and "xyz.com".
How can I do for this?
My system is sa 3.1.5.
Thanks for your time.
I also submit any such "challenge/response" spams I receive at home to
spamcop.
Eventually someone will get a clue.
Phil
-Original Message-
From: Matt Kettler [mailto:[EMAIL PROTECTED]
Sent: Friday, January 19, 2007 5:06 AM
To: dan li
Cc: users@spamassassin.apache.org
Subject: Re: box t
39 matches
Mail list logo
