Re: [exim] smtp transport process returned non-zero status 0…

On Sat, 2004-09-11 at 03:24, Philip Hazel wrote:
> On Fri, 10 Sep 2004, Jesse Regier wrote:
>
> > > > T=remote_smtp defer (-1): smtp transport process returned non-zero
> > > > status 0x0009: terminated by signal 9
>
> > After some more investigation I noticed that this was only happening
> > when a spam was being redirected to a different address. I was using
> > the following without the redirect_router line. After I added that line
> > then things started working fine. I'm wondering if there was some sort
> > of loop happening and the kernel was killing it off.
>
> A straighforward loop seems unlikely, because if there was a loop, it
> would be during the routing process. The crash you report above is in a
> *transport* process, which doesn't get started until all the routing is
> over.

I found these with debugging turned on. (/usr/exim/bin/exim -bdf
-d+all -q15m)


15:05:16 9044 set_process_info: 9044 reading output from
|/usr/exim/bin/exim -d -oMr spam-scanned -bS

15:05:25 9044 Too much output from pipe - killed


Could this have anything to do with it? It would seem to me that these
would happen well before it got to the remote_smtp transport but it's
the only place I can see anything getting killed. Would "use_bsmtp" in
the spamcheck transport show an error like this?

Will the above debugging show everything after the pipe transport to
check spam? There is quite a maze of processes going on when calling
the spamcheck transport.


spamcheck:
driver = pipe
command = /usr/exim/bin/exim -d -oMr spam-scanned -bS
use_bsmtp = true

transport_filter = /usr/bin/spamc -u "${lookup mysql{select username
from users where
username=${quote:$local_part@$domain}}{$value}{default}}"

home_directory = "/tmp"
current_directory = "/tmp"
user = mail
group = mail
log_output = true
return_fail_output = true
return_path_add = false
message_prefix =
message_suffix =


Thanks,


--
Jesse Regier <jregier@???>
Computer Systems, Inc.