Postfix playing up

I was receiving emails fine for all clients then 2 days ago one of them stopped receiving mail and I could see nothing in the logs.

I restarted postfix but it wouldn’t start:

Failed to start service: /usr/sbin/postconf: warning: /etc/postfix/ undefined parameter: mua_sender_restrictions
/usr/sbin/postconf: warning: /etc/postfix/ undefined parameter: mua_client_restrictions
/usr/sbin/postconf: warning: /etc/postfix/ undefined parameter: mua_helo_restrictions

Repeated over and over again. I’ve searched around and found an earlier post ( which says Virtualmin doesn’t install mua_ stuff and to remove them from the file. When I do that postfix goes fatal:

Sep 22 16:29:14 sunflower postfix/postfix-script[6795]: starting the Postfix mail system
Sep 22 16:29:14 sunflower postfix/master[6797]: fatal: /etc/postfix/ line 10: bad transport type: master
Sep 22 16:29:15 sunflower postfix/master[6796]: fatal: daemon initialization failure
Sep 22 16:29:16 sunflower postfix/postfix-script[6802]: fatal: mail system startup failed

According to Virtualmin the Postfix server is not working but strangely enough I am getting the occasional email slipping through.

Any ideas how to fix this?

Edit these parts get them out so everything with mua_helo_restrictions
/etc/postfix/ undefined parameter: mua_helo_restrictions

If you edit wrong so some (hidden) false chars are on the lines it could go wrong.
Then better put the important lines one after eachother in and comment these false lines out, this way i sorted some kind of same problem out.

Then it should go bettter had this also a few times but don’t know with wich setting or point these were set bij virtualmin.

Take care of updating the virtualmin config package.

You can post her with wich install script version and versions of virtualmin, fresh install or update for jamie/joe support.
They can then tell you i hope if you need to > virtualmin conifg-system … postfix

These are only WARNINGS. Are you using secure SASL/SMTP? If you are then there is a bug in WebMin when saving the configuration file. It’s only a single character bug but it prevents Postfix from working. You need to manually edit your file and where it says:

smtp_sasl_auth_enable = yes

it should in fact say

smtpd_sasl_auth_enable = yes <— NOTE THE EXTRA D

Of course, if you’re not using SASL/SMTP then it will be something totally different.

Thanks for the replies. Seems the was missing all the IP addresses. This happened after a forced refresh of the tables.

@jfro - thanks, there’s definitely something there. We’ve got postfix going now we’ve readded the IP addresses but a postfix check still brings up those warnings… what are they? Do we need them (ie should we remove them or try to fix them?)

@adelphia - checked both the and and spelling is correct for smtpd throughout but that sounds like one of those that could have you pulling your hair out over :slight_smile:

Somewhere in the forum i also has seen only warning but that they shouldn’t be added by virtualmin sofar ik know, if someone could find out with wich setting they are set could be helpfull, i did to much at a time so don’t know,

I don’t wanted these warnings so first commented them out, an put the right rules beneath, after nog error / warning i deleted them.

Is this the best i don’t know.

That part is/was also hmm not neathly inserted, (not nice linebreaks and so on) so you get if editing very fast a error there.

I don’t think that’s a bug. There are two options: One for server-side behavior (smtpd) for when Postfix is receiving mail and one for client-side behavior (smtp) when Postfix is sending mail. Webmin has options for both. If the UI is misleading (and it might be), it might be worth changing the language on that form, but we definitely want both options in there.

All I know is that after I’ve configured my mail server from within WebMin, if I do not change the smtp to smtpd, my emails are always stuck in the queue. Until I edit it to the smtpd option and then restart postfix I can not send, but can receive emails; my outgoing queue keeps getting bigger and bigger. Perhaps it’s something else that I have set incorrectly, but it works the way I’m doing it so “if it ain’t broke, don’t fix it”, if you know what I mean :wink: