Jo Rhett wrote:
I'm kindof hoping that there will be some way to get SA to re-read the
rules *WITHOUT* restarting the process.
Jonas Eckerman wrote:
Tell the daemon (or whatever) to reload the filter. The daemon creates
one or more new SA object without closing it's listening
socket/port/what
Jo Rhett wrote:
I'm kindof hoping that there will be some way to get SA to re-read the
rules *WITHOUT* restarting the process.
If your software that does it's own dameonization of SA by instantiating an
object (such as MIMEDefang or IIRC Amavis), it should be possible. Something
like this sh
Nigel Frankcom wrote:
My point, if not particularly well elucidated, is that individual
problems with MTA implementations are the realm of the particular MTA
author/s. Myself and many, many others have no issues with
ALL_TRUSTED. This issue seems to be one that's limited to Amavis, a
server that
Please reply only to the list. There is no need to CC me since I get
the post from the SA list.
My point, if not particularly well elucidated, is that individual
problems with MTA implementations are the realm of the particular MTA
author/s. Myself and many, many others have no issues with
ALL_TRU
Nigel Frankcom wrote:
On Thu, 19 Oct 2006 01:18:18 -0700, Jo Rhett
<[EMAIL PROTECTED]> wrote:
And as I've stated several times before, spamassassin *DOES* run.
Always. It's just whether or not it's doing anything useful. When it
can't talk to the sockets, it's dead in the water.
Frank Bur
On Thu, 19 Oct 2006 01:18:18 -0700, Jo Rhett
<[EMAIL PROTECTED]> wrote:
>>> And as I've stated several times before, spamassassin *DOES* run.
>>> Always. It's just whether or not it's doing anything useful. When it
>>> can't talk to the sockets, it's dead in the water.
>
>Frank Bures wrote:
And as I've stated several times before, spamassassin *DOES* run.
Always. It's just whether or not it's doing anything useful. When it
can't talk to the sockets, it's dead in the water.
Frank Bures wrote:
Interesting. Never came across that one. In my case if the socket is busy,
spamd di
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Tue, 17 Oct 2006 11:03:20 -0700, Jo Rhett wrote:
>Frank Bures wrote:
>> Or you can check that spamassassin is running after restart and if not,
start
>> it again. Also you can check that there actually was an update before
doing
>> the restart
Frank Bures wrote:
Or you can check that spamassassin is running after restart and if not, start
it again. Also you can check that there actually was an update before doing
the restart in the first place. Works for me :-)
I do the latter already.
And as I've stated several times before, spa
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Mon, 16 Oct 2006 21:56:36 -0400, Daryl C. W. O'Shea wrote:
>Jo Rhett wrote:
>> Daryl C. W. O'Shea wrote:
>>> For now, running an sa-update, then a spamassassin --lint, and then
>>> restarting is pretty safe though.
>>> sa-update [whatever] && spam
Daryl C. W. O'Shea wrote:
eh... The socket issues experienced during restarts have nothing to do
with sa-update.
Other than that sa-update is the only reason I have to restart that
process ... :-)
No, it's been more than the script which stops the process has
returned, but when the startup
Jo Rhett wrote:
Daryl C. W. O'Shea wrote:
I've been having some issues with the restarts, and when that happens
mail is down.
I'd say that's an issue of it's own, regardless of what sa-update does.
eh?
eh... The socket issues experienced during restarts have nothing to do
with sa-update
Daryl C. W. O'Shea wrote:
sa-update already lints the updates themselves (at least what it can
without loading any plugins). It'd be awfully silly to have it
installing updates that it didn't think were any good.
That's good to know.
I've been having some issues with the restarts, and when t
Jo Rhett wrote:
Daryl C. W. O'Shea wrote:
I overlooked a couple ways that you could legitimately break a --lint
between updates, nobody else caught it either, so sa-update was (the
only thing) broken for some people in the SA 3.1.6 release. SA 3.1.7
reverted to the simpler --lint check of the
I agree that it would be smart for SA to have a "graceful" restart
functionality. But wouldn't it be more responsible for the rule-set
distributors to simply test rule-sets before releasing them? Or is it a
combination of rule-sets that can break things and testing all possibilities is
not poss
Daryl C. W. O'Shea wrote:
I overlooked a couple ways that you could legitimately break a --lint
between updates, nobody else caught it either, so sa-update was (the
only thing) broken for some people in the SA 3.1.6 release. SA 3.1.7
reverted to the simpler --lint check of the update itself on
Jo Rhett wrote:
Daryl C. W. O'Shea wrote:
> this case, it's just a coincidence that I happen to provide the SARE
> sa-update channel infrastructure too.
Well Thank You for that. I love sa-update.
Thank Theo, he's behind sa-update. I like it too, which is why I'm
willing to support the SAR
17 matches
Mail list logo
