issue with mailing list delivery when using external SMTP server

[james]

Hello @msbt
I need to understand the set up first to reproduce it.

1. domain in Cloudron e.g.: cloudron.dev
2. mail settings for cloudron.dev smart host configured to use postmark
3. 2x mailboxes on Cloudron for group@cloudron.dev and team@cloudron.dev
   3a. (group|team)@cloudron.dev are configured to forward to e.g.: someinbox@gmail.com
4. a mail is sent to (group|team)@cloudron.dev and 2x cc e.g.: cc1@gmail.com and cc2@outlook.com

Resulting in:

• cc1 and cc2 delivered
• group and team saved locally on Cloudron, but never forwarded to someinbox@gmail.com

Did I get this right?

[james]

Hello @msbt
I have tested the following set up on Cloudron 10.0.4:

• domain cloudron.dev smart host configured for postmark
• local inbox groupone@cloudron.dev owned by a group named groupone
• local inbox grouptwo@cloudron.dev owned by a group named grouptwo

Forward settings:

• groupone@cloudron.dev forward settings to grouptwo@cloudron.dev and external@gmail.com
• grouptwo@cloudron.dev forward settings to external@outlook.com

Sending a mail to groupone@cloudron.dev is delivered to:

• groupone@cloudron.dev
• grouptwo@cloudron.dev
• external@gmail.com
• external@outlook.com

From this set up I was not able to reproduce your issue.

[msbt]

@james thanks for testing! In my setup the groupone@ is a regular user with 40 forwarding addresses and grouptwo@ is a shared mailbox without any forwards, but other than that you're right.

So the only difference is that groupone is owned by a single user, not a group and grouptwo is a shared mailbox with no forwards.

I can't imagine how single user vs group owners would make a difference. Did you chance anything else, like FROM address validation? Keep a copy in the inbox? Restrict senders?

[james]

Hello @msbt

All default.

[msbt]

@james can you try again without the "Keep a copy in the inbox", that's the only difference I'm seeing

edit: and maybe remove grouptwo from forwarding as well, but again, no idea how that could make a difference

[james]

Hello @msbt
No difference. Still working just fine.

[girish]

@msbt said:

@james thanks for testing! In my setup the groupone@ is a regular user with 40 forwarding addresses and grouptwo@ is a shared mailbox without any forwards, but other than that you're right.

This stood out. There is a cap of 64 forwards in the code. Could you be hitting this limit? Specifically sieve_max_redirects is 64 . I thought this was a very high number, may be not?

[msbt]

@girish unsure, it really is 40. Would there be something in a log and if so, where?

[girish]

@msbt the errors if any should be in /run/dovecot/dovecot.log or the main docker logs -f mail . You can also check if the mail is stuck in the mail queue at /app/data/haraka-queue (these are mails which are being "retried") .

[msbt]

@girish thanks, the queue is empty and the dovecot.log only has yesterday and todays events in there. I'll ask the people to do a test, they switched to manual bcc as a workaround atm