This is for Asif AND Thomas in regards to my previous email about MSMTP.
First, I did manage to get MSMTP working to send email out. While
outgoing mail worked, incoming via fetchmail suddenly stopped
complaining it couldn’t bind to 127.0.0.1 port 25:
name 0: connection to localhost:smtp [127.0.0.1/25] failed: Connection
refused.
Now, to get to the question Thomas posed to me as well as for
clarification:
We’re looking to deploy a second RT instance for our Service Department
(our IT Dept system might get rebuilt as I have yet to get an answer
from the distro list as to why my file attachments can now no longer be
over 1MB after a database mangling). This ticket desk needs to
communicate to clients our on the Internet. The thinking was to just
have a simple SMTP agent to send the mail out & have fetchmail poll the
server for incoming versus running a full-on mail server.
We run an Exchange server that relays out to a Postfix server, then to
the Internet.
Why is it Postfix is needed? Why can’t I have lighter weight agents do
the work instead? The problem I had with Postfix is for some reason on
my new RT system the moment I add the smtpd_recipient_restrictions and
put anything down there to restrict, Postfix stops working completely. I
want to control what email addresses leave the system.
Stephen Cena
QVII
MIS/IT Dept
850 Hudson Ave.
Rochester, NY. 14620
585-544-0450 x300
sjc@qvii.com
“Thank you for helping us help you help us all.”
–For email related issues, please contact postmaster@qvii.com
First, I did manage to get MSMTP working to send email out. While outgoing mail worked,
incoming via fetchmail suddenly stopped complaining it couldn’t bind to 127.0.0.1 port 25:
name 0: connection to localhost:smtp [127.0.0.1/25] failed: Connection refused.
You turned off postfix, which listened on localhost:smtp.
If you want to inject into a local MTA from fetchmail, then you need
to have a local MTA. If you don’t want a local MTA then you need to
configure fetchmail to run rt-mailgate directly or via a delivery
agent such as procmail.
We’re looking to deploy a second RT instance for our Service Department (our IT Dept system
might get rebuilt as I have yet to get an answer from the distro list as to why my file
attachments can now no longer be over 1MB after a database mangling).
Are you referring to this question, which was answered last week?
http://lists.bestpractical.com/pipermail/rt-users/2013-May/080290.html
This ticket desk needs to communicate to clients our on the
Internet. The thinking was to just have a simple SMTP agent to send
the mail out & have fetchmail poll the server for incoming versus
running a full-on mail server.
Postfix is fantastic for this. It runs trivially in a relay-only-mode
and is much more widely tested and supported than something like
msmtp.
We run an Exchange server that relays out to a Postfix server, then to the Internet.
Why is it Postfix is needed? Why can’t I have lighter weight agents do the work instead? The
problem I had with Postfix is for some reason on my new RT system the moment I add the
smtpd_recipient_restrictions and put anything down there to restrict, Postfix stops working
completely. I want to control what email addresses leave the system.
So, there was a postfix configuration problem and you’ve now spent how
many hours trying to set up msmtp instead?
-kevin