On 01/27/15 18:57, Seth wrote:
I administer an email system which uses a VPS running OpenSMTPD as the
public facing bit.
The VPS relays email to and from a separate OpenSMTPD mail server
which is located on premises. We'll call this the 'local' server.
The local server gets powered down every night, however this currently
causes messages to back up on the VPS or 'public' server. When the
local server comes back online in the morning, I have to log into the
public relay server and use smtpctl to manually resume the route to
the local server. Then a 'smtpctl schedule all' command must be run
after which the backed up overnight email comes pouring in.
This configuration is suboptimal for two reasons
* It generates bounce and 'sending delayed' postmaster messages when
the local server is down
smtpd.conf(5)
*bounce-warn* /n/{*s*|*m*|*h*|*d*}[, /.../]
Specify the delays for which temporary failure reports must be
generated when messages are stuck in the queue. For example:
bounce-warn 1h, 6h, 2d
will generate a failure report when an envelope is in the queue for
more than one hour, six hours and two days. The default is 4h.
* It requires manual intervention to ensure speedy email delivery to
the local server when it's powered back on.
I've been thinking about adding another OpenSMTPD relay mail server at
the local site, which is low power and can stay running all the time
without issue. But this merely shifts the location of the mail pile-up
from remote to local.
Any mail gurus out there have a solid method for solving this problem?
