ph10 at cus
Mar 9, 2005, 3:14 AM
Post #7 of 9
On Mon, 7 Mar 2005, Hendrik Visage wrote:
Re: Problems with multiple smtp recipients
[In reply to]
> > Please try without -v (which just gives very limited output, and has
> > overridden -d). Sorry, I should have said to use -d.
Thanks for the test. Sadly, it doesn't tell me anything.
> LOG: MAIN
> SMTP error from remote mailer after RCPT TO:<2222222 [at] sun>: host mail1.sun.ac.za [126.96.36.199]: 451 Try this a
> ddress in the next batch, please.
> reading pipe for subprocess 24170 (not ended)
> read() yielded 0
> remote delivery process 24170 ended: status=0100
I was hoping for some more debugging output from the delivery process.
Also, you didn't send ALL the debugging output. I know it is long, but
while trying to analyse this, I found myself wanting to look further
back. One of the problems is that there are multiple processes involved
here. Next time, please use -d+pid so that each debug line has a process
id added. That will make it easier to see which process is outputting what.
However, looking at the code, I regret that I can't see at the moment
how the subprocess could return that status and that debugging output.
There doesn't seem to be anything odd about your configuration, EXCEPT,
wait! I've seen something odd. You have
user = mail
in your smtp transport. Is "mail" your "exim" user? Ah, no, it isn't! I
see elsewhere in the config that you have
admin_groups = exim
exim_user = root
The last of these is worrying. It is not a good idea to run Exim as
root. And it is also less than usual to set the user at runtime. There
can be problems if you run remote transports as anything other than the
Exim user. The manual says, of the "user" option in transports:
For remote transports, you should leave this option unset unless you
really are sure you know what you are doing. When a remote transport is
running, it needs to be able to access Exim's hints databases, because
each host may have its own retry data.
I'm guessing now that the problem is caused because of user id
conflicts. There might be evidence of this in the complete debugging
Philip Hazel University of Cambridge Computing Service,
ph10 [at] cus Cambridge, England. Phone: +44 1223 334714.
Get the Exim 4 book: http://www.uit.co.uk/exim-book