On 10/02/2015 6:31 PM, Viktor Dukhovni wrote:
On Tue, Feb 10, 2015 at 06:25:22PM +1100, Carl Brewer wrote:
Have you checked that the "postmapped" virtual alias table actually
contains the address that fails verification? Report the output of:
bash-4.3# export PATH=/usr/pkg/sbin:$PATH
bash-4.3# addr="[email protected]"
bash-4.3# postmap -q "$addr" $(postconf -xh virtual_alias_maps | tr ',' ' ')
[email protected]
so it's doing the right thing
ps -o pid,etime,args -p $(pgrep -x master)
PID ELAPSED COMMAND
25095 37:15 /usr/pkg/libexec/postfix/master -w
That was started rather recently (2.11.3 upgrade). Has the problem
been observed in the last 37 minutes?
No.
Well, perhaps your system's /etc/rc scripts start the base system
Postfix. While you start the pkgsrc one when you start it by hand
(e.g. when upgrading).
One way to explain your logs is if virtual alias expansion is
suppressed somehow for verification probes, but your configuration
shows no evidence of that. The only other possibility is that the
(postmapped) table does not actually contain the address in question.
It seems to be working fine now.
thank you Viktor
Well, you have not changed anything relevant. What happens with
# /etc/rc.d/postfix stop
# /etc/rc.d/postfix start
Which master is running then? Did you edit /etc/rc.d/postfix, or
disable and create a new rc script to start the pkgsrc version?
I have:
$ grep /usr/pkg /etc/rc.d/postfix
postfix_command="/usr/pkg/sbin/${name}"
: ${required_files=`/usr/pkg/sbin/postconf -h config_directory`/main.cf}
: ${spooletcdir=`/usr/pkg/sbin/postconf -h queue_directory`/etc}
I fixed that when I originally installed the pkgsrc postfix ages ago :
grep /usr/pkg /etc/rc.d/postfix
postfix_command="/usr/pkg/sbin/${name}"
: ${required_files=`/usr/pkg/sbin/postconf -h config_directory`/main.cf}
: ${spooletcdir=`/usr/pkg/sbin/postconf -h queue_directory`/etc}
