We have been seeing a lot of reports lately of issues with alert, FAX, and VM emails not being able to send properly using SMTP via postfix with certain 3rd party relays like GMail/GSuite, Office365, 1and1, etc. This guide is intended to help you address some of these issue and get emails sending properly. You can quickly see that messages are failing if you see a 🔥 icon next to the Mail Queue item on the Dashboard. Upon hovering your mouse pointer over the 🔥 icon, you are informed that there are undelivered messages in queue:
NOTE: If the message you see upon hovering your pointer is instead a message stating that the service is not running, can't be started, or the status is unknown, you are likely instead suffering from the postfix permissions bug. You can remedy this by running:
Running this will have no ill effects on your instance, even if you aren't suffering postfix service runtime issues. So, running this is a safe first troubleshooting step. Then proceed to the specific cases below.
Pain Point 1 : Large volumes of emails via GMail/GSuite
The default GMail/GSuite SMTP server utilized by the handy GMail provider button in the System Admin > Email Setup page is the standard low-volume server that only allows up to 2000 emails sent per 24 hour period per GMail/GSuite account (NOT ALIAS). If you are using GSuite and are expecting your sent email volume to exceed 2000 messages per day, you will need to utilize the Other provider, manually specify smtp-relay.gmail.com, Use Authentication, Use TLS, & Require Both SASL options. You may need to alter your GSuite permissions to allow use of the high-volume relay.
Pain Point 2 : IPv6 doesn't work with all providers and postfix sometimes fails to switch to IPv4 properly
This seems to happen most often with Office365 but can happen with any provider. You will need to connect via SSH, run nano /etc/postfix/main.cf and find the parameter inet_protocols to change the value from all to ipv4 . This will force postfix to only send mail via IPv4.
Pain Point 3 : Your email domain settings are not being respected, resulting in SASL Authentication errors with the provider
This appears to be caused by a yet undetermined bug in either postfix or the FreePBX routines calling the postfix service. There are numerous articles discussing this issue in the various FreePBX forums. If you are seeing messages bearing an incorrect email address as the sender with a SASL error when viewing stuck messages in the mail queue (by running the mailq command), you may need to connect via SSH and complete the following steps:
Edit the following file using nano:
Go to the very end of the file (you can usually use the PgDn key to get there quickly)
You'll want to add all the lines you see below AFTER you replace "emailfromaddress" and "real-domain.com" with your correct email/authentication address:
root firstname.lastname@example.org root@localhost email@example.com firstname.lastname@example.org email@example.com root@freepbx firstname.lastname@example.org email@example.com firstname.lastname@example.org asterisk email@example.com asterisk@localhost firstname.lastname@example.org email@example.com firstname.lastname@example.org asterisk@freepbx email@example.com firstname.lastname@example.org email@example.com vm@asterisk firstname.lastname@example.org
[OPTIONAL] To rewrite the domain for ALL addresses from the PBX, use these at the bottom of the file, again replacing "real-domain.com" with your own domain:
@freepbx @real-domain @freepbx.localdomain @real-domain.com @localhost @real-domain.com @localhost.localdomain @real-domain.com
Edit the following file using nano:
Add the following line to the end:
smtp_generic_maps = hash:/etc/postfix/generic
postmap /etc/postfix/generic service postfix restart
We will update this FAQ as more information and tips become available.