| I have a new Centos 6.1 server on AWS EC2 w/ Exim 4.8.1 that I just set up
| that won't send messages. Everything seems to be configured correctly, but
| then I get the following messages in the log when I try to send anything:
|
| 2013-06-06 01:30:02 1UkGQN-0002ok-5C mta5.am0.yahoodns.net [98.136.216.25]
| Connection timed out
[...]
| I've disabled the firewall, made sure DNS resolves correctly, made sure
| Amazon has lifted the SMTP limitations, made sure that the IP isn't on any
| black lists, and that reverse DNS is correct. The HELO string in the config
| file is correct and I can telnet the remote SMTP servers on port 25 and am
| connected ok;
|
| # telnet mta5.am0.yahoodns.net 25
| Trying 98.136.216.25...
| Connected to mta5.am0.yahoodns.net.
| Escape character is '^]'.
| 220 mta1148.mail.gq1.yahoo.com ESMTP YSmtpProxy service ready
Any time there are mysterious failures in a daemon where things work
from the command line on a Linux machine, I suspect SELinux. Do you
happen to have SELinux enabled? If so, is it logging anything that you
can see? SELinux logs go to /var/log/audit/* instead of to syslog, if I
remember right.
If this is SELinux you have two options: wrestling SELinux into shape
in various ways or turning it off entirely. Turning it off entirely is
common, although it makes the SELinux people sad and unhappy.
- cks
