Receiving mails only after restarting containers

dusancoko

Hello,
I’ve looked around for a solution to this issue but couldn’t find anything. I apologize if it was already answered.
So I run a Linode instance for mailcow server, and after reading tutorials, helpful guides and posts from community, I got everything setup.
Containers are running as seen on the image:
Mailcow containers
And the DNS shows everything correctly.
Domain DNS

I can send emails and they are valid, but the problem is with receiving them. At first, I searched that I couldn’t receive emails as an issue and there were MX problems as the main culprit, but with the configuration shown, I was lost. Then I thought it was inbound rules, maybe Linode blocking ports (though I have permission for mail ports opened for the account), so I sent them a support ticket. While I waited for the response I updated my Linode instance and rebooted the machine.
Then when I logged in once again to my malcow mail I saw the messages I previously sent from my gmail account to mailcow mail. I thought it might have been DNS cache and hoped that this was solved. But then I tried to sending emails again from gmail and I wasn’t receiving anything. Then I restarted docker instances and I received new mails again.

So my question is, does anyone know what could be causing this? I don’t want to restart containers to receive mail every time.

Also if you need additional logs or anything, please let me know so I can provide them. I looked into dovecot ones, but I didn’t see anything that could implicate where the problem might be.

DocFraggle

Have a look at the postfix logs:

https://docs.mailcow.email/troubleshooting/debug-logs/

You should see if and when the sent mail is received by your server

dusancoko

DocFraggle

Hey, so I’ve checked the postfix log, but haven’t found anything regarding issues with received mails. There were some TLS services that caused warnings, but I have no idea what this is about.
TLS
But as I see in logs other IP addresses I can only assume that there are now bot scanners and such testing ports, so maybe this is a part of that.

As for the issue itself, it got resolved, but I don’t know what resolved it 😅 I think it had to be something with domain records not being recognized by other mail services even though Mailcow did recognize them. At first, it was during the restarts, but then two of the test messages I sent came an hour late without any restart. And then when I got home a few hours later, as I was preparing to reply to your response, I decided to send a message so I could track what Postfix would log, and then it came instantly. I thought it was a fluke, so I sent another, with the same result.
Then I asked the colleagues to send emails from their accounts just to check if it was only my account and emails now arrive as expected with up to 60-second delay.
And also I did receive two emails from Google dmarc with some dump files.

So I’ll keep an eye on this but as far as the issue it has been resolved. Thank you for replying!

DocFraggle

Did you check the rspamd log if it was just the greylisting feature?

https://rspamd.com/doc/modules/greylisting.html