You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by li...@arielbusiness.com on 2004/03/11 08:12:32 UTC
[N00b alert]SA in a cpanel environment
Hi,
Excuse the n00b questions, but I cannot find the answers to these
questions anywhere and have tried various cpanel forums.
I run small hosting business on a red hat box with cpanel/whm and cpanel
tend to muck around with things a bit so sometimes advice that appplies to
a "normal" install of an app does not apply to the cpanelled version :/
I would like to do the following:
1. Turn auto learn on (unless this is not recommended???).
2. Allow my clients to run sa-learn if they wish (they do not have ssh
access, but can run cron jobs), ideallty on a regular automated basis
3. Run sa-learn on a regular, automated basis for the entire server
Regarding 3. above, I have logged on as root and successfully run sa-learn
against a couple of spam and ham folders......but I think this may have
been for the root user rather than for the server as a whole.
I just did something like:
sa-learn --showdots --spam --mbox <path to mailbox file>
without specifying any paths to config files or any such thing like that,
and certainly I haven't noticed a drop-off in spam, which is why I think
it was for the root account that this was run rather than for the whole
server.
Also, ideally, once sa-learn is finished, I would like it to empty the
spam and ham folders.
thank you kindly in advance for any help you can give,
Tony
Re: Re[4]: [N00b alert]SA in a cpanel environment
Posted by Jim Conner <jc...@lrn.com>.
On Wed, 2004-03-17 at 06:52, lists@arielbusiness.com wrote:
> > A better way to zero out a file is by using '>' notation which will not
> > destroy any file descriptors attached to the file aside from being
> > extremely easy and short.
> >
> > for f in $(find . -name "whatever")
> > do
> > > $f
> > more processing on $f
> > done
>
> Hey Jim thanks for that (though presumably the processing should be done
> on $f BEFORE zeroing it out? ;)
>
:) I love *nix too. Yeah, I apologize for the obfu example. My
example was just showing a fast way to clear out files in a loop.
However, specific to your situation, you will want to massage my idea to
work best with your implementation.
- Jim
> I just love this *nix world - all these obscure abbreviations for things!
>
> cheers,
>
> Tony
--
Jim Conner <jc...@lrn.com>
Re: Re[4]: [N00b alert]SA in a cpanel environment
Posted by li...@arielbusiness.com.
> A better way to zero out a file is by using '>' notation which will not
> destroy any file descriptors attached to the file aside from being
> extremely easy and short.
>
> for f in $(find . -name "whatever")
> do
> > $f
> more processing on $f
> done
Hey Jim thanks for that (though presumably the processing should be done
on $f BEFORE zeroing it out? ;)
I just love this *nix world - all these obscure abbreviations for things!
cheers,
Tony
Re: Re[4]: [N00b alert]SA in a cpanel environment
Posted by Jim Conner <jc...@lrn.com>.
On Tue, 2004-03-16 at 23:14, lists@arielbusiness.com wrote:
> > TB> Hey cool, done that now. Just looked at the headers of a message
> > TB> received which says "autolearn=ham" This was a message from the SA
> > TB> group funnily enough - presumably that is correct?
> >
> > Unless that message included spam samples, then no problem.
> >
> > I suggest you set your non-spam auto-learn threshold to -0.01 to make
> > sure that spam that hits no rules is not accidentally learned as ham.
>
> errr...what? who? where? how???
> Also, loooking at more headers - if they say "autolearn=no", does that
> just mean SA had no idea if it was spam or ham, or does it just mean that
> autolearn is off and I Was looking at an old message? ;)else?
>
> > My understanding is that each domain with a $HOME will have one
> > $HOME/.spamassassin directory, and the bayes database built there will
> > apply to all users@domain for that domain.
>
> Cool. That does indeed seem to be the case - my mailboxes were
> refreshingly free of spam this morning - hurrah!
>
> > cp /dev/null $file
> > or
> > cat </dev/null >$file
> > are two methods I've used to empty files.
>
> okay will do that - is there any advantage of one over the other apart
> from less typing? ;)
>
A better way to zero out a file is by using '>' notation which will not
destroy any file descriptors attached to the file aside from being
extremely easy and short.
for f in $(find . -name "whatever")
do
> $f
more processing on $f
done
> > TB> - is my first ever shell script!:
> [..]
> > TB> Any obvious flaws there guys, or something I could do better? It
> >
> > Looks good to me. I wouldn't cat them all into one file first, since my
> > understanding is that the shorter/quicker sa-learn runs are better (less
> > chance they'll block bayes update by incoming email and auto-learn).
>
> okay, thanks m8. You cat them in your script though don't you?
>
> > TB> If the former, then presumably my script would be better off
> > contatenating
> > TB> the spam and ham files before passing them to a single run of
> > sa-learn?
> >
> > I run my scripts once an hour.
>
> Blimey - do you get THAT much spam? ;)
>
> > You need 200+ spams and 200+ hams before Bayes takes effect and starts
> > applying its scores to your emails. It then remains effective unless you
> > drop below those numbers (such as by deleting the database files and
> > starting over). That has nothing to do with sa-learn. The more often
> > sa-learn runs, the more current your bayes database is.
>
> Okay, thanks.
> For ham, do you just copy everything from your inbox (apart from spam not
> caught) or is there stuff you WOULDN'T put through the spam filter? eg,
> all the posts to this list?
> I am on a number of lists and the volume would make perfect ham material,
> but I'd be worried sometimes that the content wouldn't and certain
> characteristics - eg being sent to a large no. of people, me not being
> explicitly set as a recipient.
> For spam, is there any value in passing already identified spam (sent to
> the spambox thru sa-learn?
>
> thanks again,
>
> Tony
--
Jim Conner <jc...@lrn.com>
Re: Re[4]: [N00b alert]SA in a cpanel environment
Posted by li...@arielbusiness.com.
> TB> Hey cool, done that now. Just looked at the headers of a message
> TB> received which says "autolearn=ham" This was a message from the SA
> TB> group funnily enough - presumably that is correct?
>
> Unless that message included spam samples, then no problem.
>
> I suggest you set your non-spam auto-learn threshold to -0.01 to make
> sure that spam that hits no rules is not accidentally learned as ham.
errr...what? who? where? how???
Also, loooking at more headers - if they say "autolearn=no", does that
just mean SA had no idea if it was spam or ham, or does it just mean that
autolearn is off and I Was looking at an old message? ;)else?
> My understanding is that each domain with a $HOME will have one
> $HOME/.spamassassin directory, and the bayes database built there will
> apply to all users@domain for that domain.
Cool. That does indeed seem to be the case - my mailboxes were
refreshingly free of spam this morning - hurrah!
> cp /dev/null $file
> or
> cat </dev/null >$file
> are two methods I've used to empty files.
okay will do that - is there any advantage of one over the other apart
from less typing? ;)
> TB> - is my first ever shell script!:
[..]
> TB> Any obvious flaws there guys, or something I could do better? It
>
> Looks good to me. I wouldn't cat them all into one file first, since my
> understanding is that the shorter/quicker sa-learn runs are better (less
> chance they'll block bayes update by incoming email and auto-learn).
okay, thanks m8. You cat them in your script though don't you?
> TB> If the former, then presumably my script would be better off
> contatenating
> TB> the spam and ham files before passing them to a single run of
> sa-learn?
>
> I run my scripts once an hour.
Blimey - do you get THAT much spam? ;)
> You need 200+ spams and 200+ hams before Bayes takes effect and starts
> applying its scores to your emails. It then remains effective unless you
> drop below those numbers (such as by deleting the database files and
> starting over). That has nothing to do with sa-learn. The more often
> sa-learn runs, the more current your bayes database is.
Okay, thanks.
For ham, do you just copy everything from your inbox (apart from spam not
caught) or is there stuff you WOULDN'T put through the spam filter? eg,
all the posts to this list?
I am on a number of lists and the volume would make perfect ham material,
but I'd be worried sometimes that the content wouldn't and certain
characteristics - eg being sent to a large no. of people, me not being
explicitly set as a recipient.
For spam, is there any value in passing already identified spam (sent to
the spambox thru sa-learn?
thanks again,
Tony
Re: Re[8]: [N00b alert]SA in a cpanel environment
Posted by li...@arielbusiness.com.
Hi Bob,
> lac> sorry, do you mean my local (to the account) config, or the config
> for SA
> lac> on the server generally?
>
> That's up to you. system-wide or user-specific depends on how you want to
> use it.
Okey doke, tnx. I will now go look up the docs to find out exactly what
it means so I can decide where I want to put it ;)
> lac> Also, I see you using yet another method of emptying the mailbox:
> lac> cat ~/mynull > sa.learn.spam
>
> lac> Any particular reason for using this one?
>
> Some email programs want specific header emails at the top of their
> folders. This is one way of providing one such initial header email.
Ahhhh.....fair nuff.
tnx again very much for your help.
Tony
Re[8]: [N00b alert]SA in a cpanel environment
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello Tony,
Wednesday, March 17, 2004, 4:57:09 AM, you wrote:
>> I use:
>>> auto_learn_threshold_nonspam -2
>>> bayes_auto_learn_threshold_nonspam -2
>> I forget which version applies to 2.5x and which to 2.6x -- adapt to the
>> score you want to use as a threshold, and put it into your local config
>> file (eg: local.cf or whatever).
lac> sorry, do you mean my local (to the account) config, or the config for SA
lac> on the server generally?
That's up to you. system-wide or user-specific depends on how you want to
use it.
lac> Sorry, yes. What's this corpus file for?
a) to provide resources with which to re-learn spam and ham after losing
a Bayes database.
b) to provide emails against which we can test new custom rules we
develop (per http://www.exit0.us/ )
c) to provide emails against which new rules created by the devs can be
run, to assist with the setting of scores for those rules.
lac> Also, I see you using yet another method of emptying the mailbox:
lac> cat ~/mynull > sa.learn.spam
lac> Any particular reason for using this one?
Some email programs want specific header emails at the top of their
folders. This is one way of providing one such initial header email.
Bob Menschel
Re: Re[6]: [N00b alert]SA in a cpanel environment
Posted by li...@arielbusiness.com.
>>> I suggest you set your non-spam auto-learn threshold to -0.01 to make
>>> sure that spam that hits no rules is not accidentally learned as ham.
>
> tac> errr...what? who? where? how???
>
> I use:
>> auto_learn_threshold_nonspam -2
>> bayes_auto_learn_threshold_nonspam -2
> I forget which version applies to 2.5x and which to 2.6x -- adapt to the
> score you want to use as a threshold, and put it into your local config
> file (eg: local.cf or whatever).
sorry, do you mean my local (to the account) config, or the config for SA
on the server generally?
> tac> Also, loooking at more headers - if they say "autolearn=no", does
[..]
> No, autolearn=no simply means the email didn't score high enough (as
> spam) or low enough (as non-spam) to be auto-learned. It means auto-learn
> is on, but the email message didn't qualify.
gotchya, tnx!>
>>> cp /dev/null $file
>>> or
>>> cat </dev/null >$file
>>> are two methods I've used to empty files.
>
> tac> okay will do that - is there any advantage of one over the other
> apart
> tac> from less typing? ;)
>
> None that we can measure.
LOL, fair nufff, just wondering why you use BOTH.
>>> Looks good to me. I wouldn't cat them all into one file first, since
> tac> okay, thanks m8. You cat them in your script though don't you?
>
> No. My commands are:
>> sa-learn --spam --mbox sa.learn.spam # do the sa-learn
>> ls -lF `pwd`/sa.learn.spam # record this file in my log
>> cat sa.learn.spam >>~/mail/cw-spam/inbox # append to my corpus
>> cat ~/mynull > sa.learn.spam # empty the mailbox
Sorry, yes. What's this corpus file for?
Also, I see you using yet another method of emptying the mailbox:
cat ~/mynull > sa.learn.spam
Any particular reason for using this one?
>>> I run my scripts once an hour.
>
> tac> Blimey - do you get THAT much spam? ;)
>
> 7-8k spam a week, and will probably hit 9k around June.
No wonder you run it so often ;)
> I do not sa-learn the SA mailing lists, nor any other mail which contains
> samples of spam, nor discussions of spam. Otherwise I sa-learn
> everything.
Thanks for that. Funnily enough, looking in my spambox, sa had been
tagging certain posts from this list as SPAM because they containted
samples of spam.....must add the list to my whitelist ;)
> b) For simplicity of management, I dump spam from all three domains into
> one spamtrap. It's easier for me then to sa-learn all of them, rather
> than keep the already learned spam separate.
good idea! You've been a great help Bob, many thanks.
Tony
Re[6]: [N00b alert]SA in a cpanel environment
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello tony,
Tuesday, March 16, 2004, 11:08:43 PM, you wrote:
>> TB> Hey cool, done that now. Just looked at the headers of a message
>> TB> received which says "autolearn=ham" This was a message from the SA
>> TB> group funnily enough - presumably that is correct?
>>
>> Unless that message included spam samples, then no problem.
>>
>> I suggest you set your non-spam auto-learn threshold to -0.01 to make
>> sure that spam that hits no rules is not accidentally learned as ham.
tac> errr...what? who? where? how???
I use:
> auto_learn_threshold_nonspam -2
> bayes_auto_learn_threshold_nonspam -2
I forget which version applies to 2.5x and which to 2.6x -- adapt to the
score you want to use as a threshold, and put it into your local config
file (eg: local.cf or whatever).
tac> Also, loooking at more headers - if they say "autolearn=no", does that
tac> just mean SA had no idea if it was spam or ham, or does it just mean that
tac> autolearn is off and I Was looking at an old message? ;)else?
No, autolearn=no simply means the email didn't score high enough (as
spam) or low enough (as non-spam) to be auto-learned. It means auto-learn
is on, but the email message didn't qualify.
>> My understanding is that each domain with a $HOME will have one
>> $HOME/.spamassassin directory, and the bayes database built there will
>> apply to all users@domain for that domain.
tac> Cool. That does indeed seem to be the case - my mailboxes were
tac> refreshingly free of spam this morning - hurrah!
>> cp /dev/null $file
>> or
>> cat </dev/null >$file
>> are two methods I've used to empty files.
tac> okay will do that - is there any advantage of one over the other apart
tac> from less typing? ;)
None that we can measure.
>> TB> - is my first ever shell script!:
tac> [..]
>> TB> Any obvious flaws there guys, or something I could do better? It
>>
>> Looks good to me. I wouldn't cat them all into one file first, since my
>> understanding is that the shorter/quicker sa-learn runs are better (less
>> chance they'll block bayes update by incoming email and auto-learn).
tac> okay, thanks m8. You cat them in your script though don't you?
No. My commands are:
> sa-learn --spam --mbox sa.learn.spam # do the sa-learn
> ls -lF `pwd`/sa.learn.spam # record this file in my log
> cat sa.learn.spam >>~/mail/cw-spam/inbox # append to my corpus
> cat ~/mynull > sa.learn.spam # empty the mailbox
>> TB> If the former, then presumably my script would be better off
>> contatenating
>> TB> the spam and ham files before passing them to a single run of
>> sa-learn?
>>
>> I run my scripts once an hour.
tac> Blimey - do you get THAT much spam? ;)
7-8k spam a week, and will probably hit 9k around June.
>> You need 200+ spams and 200+ hams before Bayes takes effect and starts
>> applying its scores to your emails. It then remains effective unless you
>> drop below those numbers (such as by deleting the database files and
>> starting over). That has nothing to do with sa-learn. The more often
>> sa-learn runs, the more current your bayes database is.
tac> Okay, thanks.
tac> For ham, do you just copy everything from your inbox (apart from spam not
tac> caught) or is there stuff you WOULDN'T put through the spam filter? eg,
tac> all the posts to this list?
I do not sa-learn the SA mailing lists, nor any other mail which contains
samples of spam, nor discussions of spam. Otherwise I sa-learn
everything.
tac> I am on a number of lists and the volume would make perfect ham material,
tac> but I'd be worried sometimes that the content wouldn't and certain
tac> characteristics - eg being sent to a large no. of people, me not being
tac> explicitly set as a recipient.
If you (or your people) ever get non-spam from the same people who use
those lists, discussing the same topics, then learning them will be a big
help (avoid FPs).
tac> For spam, is there any value in passing already identified spam (sent to
tac> the spambox thru sa-learn?
a) If you do multiple domains as I do, then it's easier to simply feed
all spam into all three domains rather than figure out where it has
already been learned.
b) For simplicity of management, I dump spam from all three domains into
one spamtrap. It's easier for me then to sa-learn all of them, rather
than keep the already learned spam separate.
Bob Menschel
Re[4]: [N00b alert]SA in a cpanel environment
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello Tony,
Tuesday, March 16, 2004, 2:45:59 PM, you wrote:
>> If you have access to the /etc/mail/spamassassin/local.cf
>> file (may be in a different directory according to how SA is
>> called), then you can add the parameter
>> > bayes_auto_learn 1
TB> Hey cool, done that now. Just looked at the headers of a message
TB> received which says "autolearn=ham" This was a message from the SA
TB> group funnily enough - presumably that is correct?
Unless that message included spam samples, then no problem.
I suggest you set your non-spam auto-learn threshold to -0.01 to make
sure that spam that hits no rules is not accidentally learned as ham.
TB> I managed to get sa-learn to work for a non-root account by deleting the
TB> bayes* files as you suggested.
TB> Presumably the bayes database applies then to any user@domain for the
TB> userid I run it under?
My understanding is that each domain with a $HOME will have one
$HOME/.spamassassin directory, and the bayes database built there will
apply to all users@domain for that domain.
TB> I've had a look at your script and it's given me some ideas thanks - I have
TB> written a script which will look for all files called learn_spam or
TB> learn_ham and run sa-learn on them, then "empties" the files by removing
TB> them and touching them (is there a better way?)
cp /dev/null $file
or
cat </dev/null >$file
are two methods I've used to empty files.
TB> I know nothing about shell programming other than what I have picked up from
TB> Bob's script and google, so forgive if it's a little rough around the edges
TB> - is my first ever shell script!:
TB> ====================================
TB> #!/bin/sh
TB> if [ $1 -eq "d" ] ; then
TB> SARGS="--showdots"
TB> fi
TB> echo "Learning SPAM"
TB> for FILE in `find $HOME -name learn_spam -print`
TB> do
TB> echo "Processing $FILE"
TB> sa-learn --spam --mbox $FILE $SARGS
TB> rm $FILE
TB> touch $FILE
TB> done
TB> echo "Learning HAM"
TB> for FILE in `find $HOME -name learn_ham -print`
TB> do
TB> echo "Processing $FILE"
TB> sa-learn --ham --mbox $FILE $SARGS
TB> rm $FILE
TB> touch $FILE
TB> done
TB> echo "Done"
TB> ====================================
TB> Any obvious flaws there guys, or something I could do better? It *seems*
TB> to work okay anyway.
TB> Should I bung them all into one file first????
Looks good to me. I wouldn't cat them all into one file first, since my
understanding is that the shorter/quicker sa-learn runs are better (less
chance they'll block bayes update by incoming email and auto-learn).
TB> The other thing is, how often should I run it - I've seen it mentioned
TB> before that you need about 200 spams and 200 hams for sa-learn to be
TB> effective - does this mean 200 _per run_ or that you need to have learned
TB> about that number in total for it to be effective?
TB> If the former, then presumably my script would be better off contatenating
TB> the spam and ham files before passing them to a single run of sa-learn?
I run my scripts once an hour.
You need 200+ spams and 200+ hams before Bayes takes effect and starts
applying its scores to your emails. It then remains effective unless you
drop below those numbers (such as by deleting the database files and
starting over). That has nothing to do with sa-learn. The more often
sa-learn runs, the more current your bayes database is.
Bob Menschel
RE: Re[2]: SA in a cpanel environment - NOW tie failed: File exists
Posted by Tony Butler <li...@arielbusiness.com>.
> If you have access to the /etc/mail/spamassassin/local.cf file (may be
> in a different directory according to how SA is called), then you can
> add the parameter
> > bayes_auto_learn 1
Hey cool, done that now. Just looked at the headers of a message received
which says "autolearn=ham"
This was a message from the SA group funnily enough - presumably that is
correct?
I managed to get sa-learn to work for a non-root account by deleting the
bayes* files as you suggested.
Presumably the bayes database applies then to any user@domain for the userid
I run it under?
I've had a look at your script and it's given me some ideas thanks - I have
written a script which will look for all files called learn_spam or
learn_ham and run sa-learn on them, then "empties" the files by removing
them and touching them (is there a better way?)
I know nothing about shell programming other than what I have picked up from
Bob's script and google, so forgive if it's a little rough around the edges
- is my first ever shell script!:
====================================
#!/bin/sh
if [ $1 -eq "d" ] ; then
SARGS="--showdots"
fi
echo "Learning SPAM"
for FILE in `find $HOME -name learn_spam -print` do
echo "Processing $FILE"
sa-learn --spam --mbox $FILE $SARGS
rm $FILE
touch $FILE
done
echo "Learning HAM"
for FILE in `find $HOME -name learn_ham -print` do
echo "Processing $FILE"
sa-learn --ham --mbox $FILE $SARGS
rm $FILE
touch $FILE
done
echo "Done"
====================================
Any obvious flaws there guys, or something I could do better? It *seems*
to work okay anyway.
Should I bung them all into one file first????
The other thing is, how often should I run it - I've seen it mentioned
before that you need about 200 spams and 200 hams for sa-learn to be
effective - does this mean 200 _per run_ or that you need to have learned
about that number in total for it to be effective?
If the former, then presumably my script would be better off contatenating
the spam and ham files before passing them to a single run of sa-learn?
> I like salt-free sunflower seeds...
They're on their way!
Cheers,
Tony
Re[2]: SA in a cpanel environment - NOW tie failed: File exists
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello Tony,
Monday, March 15, 2004, 3:02:23 PM, you wrote:
>> If you have access to the /etc/mail/spamassassin/local.cf
>> file (may be in a different directory according to how SA is
>> called), then you can add the parameter
>> > bayes_auto_learn 1
>> to that file.
TB> Okey doke, done that...now how can I tell if it worked???
TB> Do I need to restart something maybe?
If you're running spamd/spamc or equivalent, you normally need to restart
in order to have parameter changes take effect.
If auto-learn is turned on, then the bayes databases will be updated by
emails which exceed the auto-learn thresholds in either direction. I find
three bayes database files in my $home/.spamassassin directory:
bayes_journal, bayes_seen, and bayes_toks. All three are updated during
the auto-learn process.
>> T> How much do you want for a copy of your scripts? ;)
>>
>> I like salt-free sunflower seeds...
TB> LOL, send me your address and they're on their way ;)
Bob Menschel
P. O. Box 2369
Citrus Heights, CA 95611
Donations gratefully accepted from all sources. Not tax deductible.
TB> Had a look at the script and can even reasonably understand it - is not as
TB> complex as I expected - hurrah!
TB> That's the good news.
TB> The bad news is, when I try (not using your script _just_ yet) to run
TB> sa-learn NOT logged on as root (I know root works) I get an error, which
TB> tells me to run it using -D, which produces this:
TB> -bash-2.05b$ sa-learn --spam --showdots --mbox
TB> mail/arielbusiness.com/abutler/spam -D
TB> debug: Score set 0 chosen.
TB> ...
TB> debug: using "/home/arielbus/.spamassassin/user_prefs" for user prefs file
TB> debug: bayes: 30704 tie-ing to DB file R/O
TB> /home/arielbus/.spamassassin/bayes_toks
TB> Cannot open bayes databases /home/arielbus/.spamassassin/bayes_* R/O: tie
TB> failed:
TB> ...
TB> Cannot open bayes databases /home/arielbus/.spamassassin/bayes_* R/W: tie
TB> failed: File exists
TB> ...
TB> Cannot open bayes databases /home/arielbus/.spamassassin/bayes_* R/W: tie
TB> failed: File exists
TB> ...
TB> Question: what does it all mean & how can I fix it???????
Looks to me like you either have a permissions problem with those bayes_*
files, or the files and software are incompatible.
When you issue the command above, what userid are you, who owns those
files, and what are their permissions? If owned by root and changable
only by root, that could cause this problem.
Is this an upgrade from an earlier (2.5x or older) version? If so, did
you follow the install instructions for updating the database?
Do you have the correct perl DB package installed at a reasonably current
level?
When my host moved my domains from 2.5x to 2.6x, two of my Bayes
databases migrated fine, and one ended up "broken" with symptoms not
dissimilar from what you show. I deleted those bayes_* files, waited for
SA to recreate them during normal processing, and then used sa-learn to
reload them with current emails. Worked fine.
Bob Menschel
Re[2]: [N00b alert]SA in a cpanel environment
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello Tony,
Thursday, March 11, 2004, 11:25:59 PM, you wrote:
>>> I would like to do the following:
>>
>>> 1. Turn auto learn on (unless this is not recommended???).
>>
>> In my cpanel installations auto-learn is automatic.
T> Any idea how I can turn this on? It is turned off on my server, and those
T> of others using the same dedicated server provider - and they don't do it
T> themselves, so I assume it is OFF by default in recent versions of
T> CPanel....
If you have access to the /etc/mail/spamassassin/local.cf file (may be in
a different directory according to how SA is called), then you can add
the parameter
> bayes_auto_learn 1
to that file. See http://www.spamassassin.org/doc/Mail_SpamAssassin_Conf.html
>>> 2. Allow my clients to run sa-learn if they wish (they do not have ssh
>>> access, but can run cron jobs), ideallty on a regular automated basis
>>
>> I use cron scripts which look for files named newspam.txt and
>> newham.txt in my domain root directories, and if found, run sa-learn
>> against and then delete the files. I create these files by FTP upload
>> from my email client. Works great.
T> Cool. Do you run these for the whole server or just for YOUR domain?
Just my domain (it's the only area I have permissions for). It would not
be hard to use similar logic to scan the entire server.
>> Only requirement is an email client that will output
>> intelligent/useful files (I use mbox format).
T> ditto.
>> I have users that use horde for webmail. I've created hamlearn and
>> spamlearn folders for them, and my cron scripts look to see if those
>> folders are non-empty. If so, it runs sa-learn against those folders
>> and then empties them.
T> Sounds good, just the sort of thing I'm after.
>>> 3. Run sa-learn on a regular, automated basis for the entire server
>>
>> That can be a challenge.
T> running it for the entire server, or running it automatically?
Running it for the entire server, and allowing people to have their own
customized database.
>> sa-learn runs best for individual users and/or individual systems.
>> While a system-wide bayes is better than none, you won't want to
>> interfere with those users/domains that are able to do their own bayes
>> work.
T> Can you tell me how to run it for the entire server?
T> Why is it better to have individual ones for each user - becuause they
T> have different ideas of what constitutes spam?
Not only different ideas of what constitutes spam, but also actually
different spam, and more importantly, different ham. The domain used by a
rape support group will have VERY different ham than the domain used by a
mortgage banker.
The way to run it for the entire server would be to run it as a specific
userid, and to set up local.cf to ensure that EVERYONE uses that specific
userid's bayes database. Others can tell you how to do that.
>> As an end-user, I can't help you much with the cpanel setup, but I can
>> help you with the types of things I have managed to do here.
T> Sounds like you're a lot more than an end-user Bob.
T> You are doing a heck of a lot that I would like to be able to do.
T> Apart from the lack of SA/cpanel knowledge, I am a *indows man not a *nix
T> one, so cron jobs/perl scripts/shell scripts are all pretty new to me.
T> How much do you want for a copy of your scripts? ;)
I like salt-free sunflower seeds...
Bob Menschel
Re: [N00b alert]SA in a cpanel environment
Posted by li...@arielbusiness.com.
Hi Bob,
> Hello lists,
You can call me Tony ;)
Many thanks for responding to my plea.
>> a "normal" install of an app does not apply to the cpanelled version :/
>
> Understood. I'm a domain master on three similar machines (three
> different domains), all using SA under cpanel's installation. Works
> well for me (though I've strengthened SA for myself more than most
> others would be able to).
Ah, a guru then ;)
>> I would like to do the following:
>
>> 1. Turn auto learn on (unless this is not recommended???).
>
> In my cpanel installations auto-learn is automatic.
Any idea how I can turn this on? It is turned off on my server, and those
of others using the same dedicated server provider - and they don't do it
themselves, so I assume it is OFF by default in recent versions of
CPanel....
>> 2. Allow my clients to run sa-learn if they wish (they do not have ssh
>> access, but can run cron jobs), ideallty on a regular automated basis
>
> I use cron scripts which look for files named newspam.txt and
> newham.txt in my domain root directories, and if found, run sa-learn
> against and then delete the files. I create these files by FTP upload
> from my email client. Works great.
Cool. Do you run these for the whole server or just for YOUR domain?
> Only requirement is an email client that will output
> intelligent/useful files (I use mbox format).
ditto.
> I have users that use horde for webmail. I've created hamlearn and
> spamlearn folders for them, and my cron scripts look to see if those
> folders are non-empty. If so, it runs sa-learn against those folders
> and then empties them.
Sounds good, just the sort of thing I'm after.
>> 3. Run sa-learn on a regular, automated basis for the entire server
>
> That can be a challenge.
running it for the entire server, or running it automatically?
>> and certainly I haven't noticed a drop-off in spam, which is why I think
>> it was for the root account that this was run rather than for the whole
>> server.
>
> I agree -- sa-learn normally runs for whatever user runs it.
> sa-learn runs best for individual users and/or individual systems.
> While a system-wide bayes is better than none, you won't want to
> interfere with those users/domains that are able to do their own bayes
> work.
Can you tell me how to run it for the entire server?
Why is it better to have individual ones for each user - becuause they
have different ideas of what constitutes spam?
> As an end-user, I can't help you much with the cpanel setup, but I can
> help you with the types of things I have managed to do here.
Sounds like you're a lot more than an end-user Bob.
You are doing a heck of a lot that I would like to be able to do.
Apart from the lack of SA/cpanel knowledge, I am a *indows man not a *nix
one, so cron jobs/perl scripts/shell scripts are all pretty new to me.
How much do you want for a copy of your scripts? ;)
many thanks,
Tony
Re: [N00b alert]SA in a cpanel environment
Posted by Bob Menschel <Ro...@Menschel.net>.
Hello lists,
Wednesday, March 10, 2004, 11:12:32 PM, you wrote:
> Hi,
> Excuse the n00b questions, but I cannot find the answers to these
> questions anywhere and have tried various cpanel forums.
> I run small hosting business on a red hat box with cpanel/whm and cpanel
> tend to muck around with things a bit so sometimes advice that appplies to
> a "normal" install of an app does not apply to the cpanelled version :/
Understood. I'm a domain master on three similar machines (three
different domains), all using SA under cpanel's installation. Works
well for me (though I've strengthened SA for myself more than most
others would be able to).
> I would like to do the following:
> 1. Turn auto learn on (unless this is not recommended???).
In my cpanel installations auto-learn is automatic.
> 2. Allow my clients to run sa-learn if they wish (they do not have ssh
> access, but can run cron jobs), ideallty on a regular automated basis
I use cron scripts which look for files named newspam.txt and
newham.txt in my domain root directories, and if found, run sa-learn
against and then delete the files. I create these files by FTP upload
from my email client. Works great.
Only requirement is an email client that will output
intelligent/useful files (I use mbox format).
I have users that use horde for webmail. I've created hamlearn and
spamlearn folders for them, and my cron scripts look to see if those
folders are non-empty. If so, it runs sa-learn against those folders
and then empties them.
> 3. Run sa-learn on a regular, automated basis for the entire server
That can be a challenge.
> Regarding 3. above, I have logged on as root and successfully run sa-learn
> against a couple of spam and ham folders......but I think this may have
> been for the root user rather than for the server as a whole.
> I just did something like:
> sa-learn --showdots --spam --mbox <path to mailbox file>
> without specifying any paths to config files or any such thing like that,
> and certainly I haven't noticed a drop-off in spam, which is why I think
> it was for the root account that this was run rather than for the whole
> server.
I agree -- sa-learn normally runs for whatever user runs it.
sa-learn runs best for individual users and/or individual systems.
While a system-wide bayes is better than none, you won't want to
interfere with those users/domains that are able to do their own bayes
work.
As an end-user, I can't help you much with the cpanel setup, but I can
help you with the types of things I have managed to do here.
Bob Menschel