Viktor Dukhovni wrote:
On Fri, May 18, 2012 at 10:11:10AM -0400, Ben Rosengart wrote:
...
2. I never liked milters (multi-threading complexity, and poor degradation
when CPU is scarce under high load).
Take a look at libmilter with a pool of workers. It's a FFR : _FFR_WORKERS_POOL
Ben Rosengart:
> On Fri, May 18, 2012 at 10:55:49AM -0400, Wietse Venema wrote:
> > Ben Rosengart:
> > > On Fri, May 18, 2012 at 02:27:11PM +, Viktor Dukhovni wrote:
> > > >
> > > > 2. I never liked milters (multi-threading complexity, and poor
> > > > degradation
> > > > when CPU is s
On Fri, May 18, 2012 at 10:55:49AM -0400, Wietse Venema wrote:
> Ben Rosengart:
> > On Fri, May 18, 2012 at 02:27:11PM +, Viktor Dukhovni wrote:
> > >
> > > 2. I never liked milters (multi-threading complexity, and poor
> > > degradation
> > > when CPU is scarce under high load).
> >
Ben Rosengart:
> On Fri, May 18, 2012 at 02:27:11PM +, Viktor Dukhovni wrote:
> >
> > 2. I never liked milters (multi-threading complexity, and poor degradation
> > when CPU is scarce under high load).
>
> If you strip away the API layer, it's a client/server protocol over a
> socket,
On Fri, May 18, 2012 at 02:27:11PM +, Viktor Dukhovni wrote:
>
> 2. I never liked milters (multi-threading complexity, and poor degradation
> when CPU is scarce under high load).
If you strip away the API layer, it's a client/server protocol over a
socket, not so different from the pol
Ben Rosengart:
> On Fri, May 18, 2012 at 06:20:04AM -0400, Wietse Venema wrote:
> > Ben Rosengart:
> > > On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
> > > > Ben Rosengart:
> > > > > I enabled verbose mode in smtpd(8), but I don't know what to look for
> > > >
> > > > Adding head
On Fri, May 18, 2012 at 10:11:10AM -0400, Ben Rosengart wrote:
> I have one last problem, which is I have code (borrowed from Viktor)
> which peeks at the files in the hold queue, and does special processing if
> X-MS-Floodstopper is present.
That code is in turn borrowed from qshape(1), which Wi
On Fri, May 18, 2012 at 06:20:04AM -0400, Wietse Venema wrote:
> Ben Rosengart:
> > On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
> > > Ben Rosengart:
> > > > I enabled verbose mode in smtpd(8), but I don't know what to look for
> > >
> > > Adding headers is implemented in cleanup
Ben Rosengart:
> On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
> > Ben Rosengart:
> > > I enabled verbose mode in smtpd(8), but I don't know what to look for
> >
> > Adding headers is implemented in cleanup(8).
>
> Before or after header_checks?
Between header_checks and milter_
On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
> Ben Rosengart:
> > I enabled verbose mode in smtpd(8), but I don't know what to look for
>
> Adding headers is implemented in cleanup(8).
Before or after header_checks?
> Headers will never be
> added when smtpd(8) is followed by a
Ben Rosengart:
> On Thu, May 17, 2012 at 02:31:19PM -0400, Wietse Venema wrote:
> >
> > I installed Sendmail::PMilter and it works out of the box (I used
> > TCP instead of UNIX-domain because I didn't want to mess with chroot
> > configuration).
>
> Do you think it makes a difference that I am ru
On Thu, May 17, 2012 at 02:31:19PM -0400, Wietse Venema wrote:
>
> I installed Sendmail::PMilter and it works out of the box (I used
> TCP instead of UNIX-domain because I didn't want to mess with chroot
> configuration).
Do you think it makes a difference that I am running this as an
smtpd_milter
Ben Rosengart:
> On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
> >
> > Adding headers is implemented in cleanup(8). Headers will never be
> > added when smtpd(8) is followed by a before-queue content filter.
> > Perhaps that is your problem?
>
> No, but I appreciate the process b
On Thu, May 17, 2012 at 01:19:52PM -0400, Wietse Venema wrote:
>
> Adding headers is implemented in cleanup(8). Headers will never be
> added when smtpd(8) is followed by a before-queue content filter.
> Perhaps that is your problem?
No, but I appreciate the process behind the guess.
> Otherwise
Ben Rosengart:
> I enabled verbose mode in smtpd(8), but I don't know what to look for
Adding headers is implemented in cleanup(8). Headers will never be
added when smtpd(8) is followed by a before-queue content filter.
Perhaps that is your problem?
Otherwise, can you provide a small Perl sample
On Thu, May 17, 2012 at 12:04:11PM -0400, Wietse Venema wrote:
> Ben Rosengart:
> > >
> > > Are you implementing your own Milter on-the-wire read/write
> > > routines?
> >
> > No, I am using Sendmail::PMilter.
>
> Is that from CPAN or elsewhere?
CPAN. It's currently unmaintained, but we are l
Ben Rosengart:
> On Thu, May 17, 2012 at 11:43:03AM -0400, Wietse Venema wrote:
> > Ben Rosengart:
> > > Dear Postfixers,
> > > I am testing a milter which, under some circumstances, adds a
> > > header to a message. The problem is that the header is simply not added.
> > >
> > > I have instrum
On Thu, May 17, 2012 at 11:43:03AM -0400, Wietse Venema wrote:
> Ben Rosengart:
> > Dear Postfixers,
> > I am testing a milter which, under some circumstances, adds a
> > header to a message. The problem is that the header is simply not added.
> >
> > I have instrumented the milter-side code an
Ben Rosengart:
> Dear Postfixers,
> I am testing a milter which, under some circumstances, adds a
> header to a message. The problem is that the header is simply not added.
>
> I have instrumented the milter-side code and I can see that the milter
> server is, at EOM time, writing to the socket
19 matches
Mail list logo
