On Apr 3, 2004, at 2:30 AM, Peter Bowyer wrote:
> Michael Johnson <kaiser@???> wrote:
>
> <snip>
>>
>> In my logs, I'm seeing lines like this:
>>
>> unable to set gid=20 or uid=501 (euid=1032): local delivery to...
>
> Check permissions on the exim binary - should be suid root:
>
> -rwsr-xr-x 1 root root 771438 Feb 25 22:28
> /usr/exim/bin/exim
>
> If it isn't, then it will still be running as the exim user when it
> comes to
> the local delivery, and so won't be able to suid to the local user.
Now exim will run without dying, but there still seems to be a problem.
I've got another exim server sitting about 2 inches from it, connected
to the same switch, UPS, and KVM even. When I try to send mail to my
sick server from the healthy one, I get log entries like:
2004-04-03 09:48:16 HVLNFA-000M45-97 == michael@???
R=dnslookup T=remote_smtp defer (-53): retry time not reached for any
host
It seems, according to the FAQ, that since my earlier attempts were
going to a misconfigured exim server, ALL the messages are in a retry
purgatory. So even though I just sent this message out, exim will not
retry until it thinks it should for all the messages to that host. Do
I understand that correctly?
With retry rules like this:
# Domain Error Retries
# ------ ----- -------
* quota
* * F,2h,15m; G,16h,1h,1.5; F,4d,6h
I could be waiting a while. Considering I started working on this over
16 hours ago, it could be an hour to 90 minutes before I see anything.
Why then would a message sent from a completely different host not make
it? I don't see any attempts made in the log. It has been 15-20
minutes since I sent the message through the foreign host (they use
Netscape). I should have seen something by now, shouldn't I?
-Michael
---------------------------------------
None are more hopelessly enslaved than those who falsely believe they
are free.
--Johann Wolfgang von Goethe.
