Hello CentOS,
Setup: CentOS 4.1 (Latest YUM) Sendmail Multiple IP's on eth0:x (1 through 150 or so) iptables
I've noticed that whenever I add another IP address to my servers on one of the eth0:x aliases, the IP comes up fine and restarting pure-ftpd and apache seems to be able to use the IP straight away. Sendmail also 'seems' to use it fine, but something odd is occuring that I have been able to track down.
When adding the IP, any domains pointing to the IP are added to the local-host-names file, their email aliases are all configured, and sendmail is restarted. The problem is, none of the aliases ever work unless I actually reboot the entire machine.
I've tried restarting as many of the running daemons as I can, I'm either looking in the wrong places, or there's no other choice? On the old RedHat machines I had, this used to work fine without a reboot. Any suggestions?
On 05/10/05, Mickael Maddison centos@silverservers.com wrote:
When adding the IP, any domains pointing to the IP are added to the local-host-names file, their email aliases are all configured, and sendmail is restarted. The problem is, none of the aliases ever work unless I actually reboot the entire machine.
This might be a stupid question but are you running newaliases after modifying /etc/aliases?
If you restart Sendmail with either
service sendmail restart|reload|stop/start /etc/rc.d/init.d/sendmail restart|reload|stop/start
the newaliases *should* be being run. But it certainly sounds like it might not be.
Will.
Hello Will,
Indeed, newaliases runs every few minutes as part of the alias gathering routines. In fact, I tested that part, and as far as the "local" server was concerned, sendmail -bv email@domain.com was set to route correctly. The problem I got was that any server connecting to the machine to send mail to the domain on the recently added IP would be rejected.
Thanks for pondering.