Andrew Kohlsmith wrote:
>>2. Use an external mail class to parse your email and split out just the
>>text and html parts to pass to SpamAssassin. A huge part of the load on
>>SA is in doing nasty slow regexps across the "rawbody" tests when the
>>email contains large MIME attachments.
>
>
> Do y
> 2. Use an external mail class to parse your email and split out just the
> text and html parts to pass to SpamAssassin. A huge part of the load on
> SA is in doing nasty slow regexps across the "rawbody" tests when the
> email contains large MIME attachments.
Do you have a document giving an ex
Justin Robinson wrote:
> Hi all, I just barely subscribed to the list and I have a question...
>
>
> We did a bit of an experiment in using SpamAssassin over the past month
> and just disabled it last night. We have roughly 9000 email accounts on
> our server and we mass-enabled spam assass
On Thu, 2002-05-16 at 12:39, Justin Robinson wrote:
> We are running FreeBSD 4.3-Release, sendmail 8.12.2, procmail 3.15.1, and
> spamassassin 2.11
Some of the bugs that have been reported and fixed since 2.11 had to do
with rules with regexps that took really long times at high loads on
some inp
We are not using Razor, so I do not believe it should be an issue with that.
Justin
-Original Message-
From: Theo Van Dinter [mailto:[EMAIL PROTECTED]]
Sent: Thursday, May 16, 2002 4:55 PM
To: Justin Robinson; spamtalk (E-mail)
Subject: Re: [SAtalk] Speed/Performance Issue
On Thu
On Thu, May 16, 2002 at 01:49:26PM -0700, Daniel Rogers wrote:
> Were you using spamc/spamd?
For a few days there, Razor was responding very slowly. Have you
optimized the SA configuration to do only local checks? With enough
waiting around for something like Razor, your load would go up
signif
On Thu, May 16, 2002 at 03:39:44PM -0400, Justin Robinson wrote:
> We did a bit of an experiment in using SpamAssassin over the past month and
> just disabled it last night. We have roughly 9000 email accounts on our
> server and we mass-enabled spam assassin for them. Only 78 used the opt-out
>
Hi all, I just
barely subscribed to the list and I have a question...
We did a bit of an
experiment in using SpamAssassin over the past month and just disabled it last
night. We have roughly 9000 email accounts on our server and we
mass-enabled spam assassin for them. Only 78 used the
>However a way around this has just occured to me:
>
>1. Run all negatively scoring tests.
>
>2. Run positively scoring tests in highest-score first order.
>
>3. Stop when we hit the threshold.
How about:
1) run the positive tests until it hits the threshold
2) run all the negative tests
3) r
On Thu, 7 Mar 2002, Bob Plankers wrote:
> I still think an instant accept would be beneficial, if it is implemented
> as a lower threshold or as an outright accept. Certainly there is
> some speed to be gained by skipping the processing altogether, but
> inititally the lower threshold would be ea
I still think an instant accept would be beneficial, if it is implemented
as a lower threshold or as an outright accept. Certainly there is
some speed to be gained by skipping the processing altogether, but
inititally the lower threshold would be easier to implement. We'd just
need another con
On Thu, 7 Mar 2002, Matt Sergeant wrote:
> On Wed, 6 Mar 2002, Bob Plankers wrote:
>
> > Two things:
> >
> > 1) You didn't implement the whitelist/blacklist outright accept/reject
> > concept yet. Bug #62 mentions that in some of Craig's notes, so if that's
> > still of interest then someone shou
On Wed, 6 Mar 2002, Bob Plankers wrote:
> Two things:
>
> 1) You didn't implement the whitelist/blacklist outright accept/reject
> concept yet. Bug #62 mentions that in some of Craig's notes, so if that's
> still of interest then someone should create a new "bug" for it.
Yeah, I'd prefer that as
Two things:
1) You didn't implement the whitelist/blacklist outright accept/reject
concept yet. Bug #62 mentions that in some of Craig's notes, so if that's
still of interest then someone should create a new "bug" for it.
2) Spamd needs an update to reflect the short-circuiting. Attached is a
p
On 5 Mar 2002, Craig Hughes wrote:
> On Tue, 2002-03-05 at 09:35, Matt Sergeant wrote:
> > On Tue, 5 Mar 2002, Yevgeniy Miretskiy wrote:
> >
> > > The question is: why do I need to run all tests if I'm running spamassassin with
>-L flag?
>
> > > Again, sorry if this topic was beaten to death bef
On Tue, Mar 05, 2002 at 11:17:09AM -0800, Craig Hughes wrote:
> On Tue, 2002-03-05 at 09:35, Matt Sergeant wrote:
> > On Tue, 5 Mar 2002, Yevgeniy Miretskiy wrote:
> >
> > > The question is: why do I need to run all tests if I'm running spamassassin with
>-L flag?
>
> > > Again, sorry if this t
On Tue, 2002-03-05 at 09:35, Matt Sergeant wrote:
> On Tue, 5 Mar 2002, Yevgeniy Miretskiy wrote:
>
> > The question is: why do I need to run all tests if I'm running spamassassin with
>-L flag?
> > Again, sorry if this topic was beaten to death before...
> The problem is that some of the scor
On Tue, 5 Mar 2002, Yevgeniy Miretskiy wrote:
> Hello everybody,
>
> I'm new to the list so please excuse me if this topic was beaten to death...
>
> I'm very impressed with spam assassin acuracy in spam detection, however
> it is not as fast as I wish it was (which is very much understandable gi
Hello everybody,
I'm new to the list so please excuse me if this topic was beaten to death...
I'm very impressed with spam assassin acuracy in spam detection, however
it is not as fast as I wish it was (which is very much understandable given
that spamassassin is written in perl and uses huge nu
19 matches
Mail list logo
