Trying to use gmail relay but does not work

[albgen]

I have tried to configure DMS as a null mailer (SMTP_ONLY=1) to relay mail through Gmail, but it is failing with 451 4.3.0 Error: queue file write error.

The username and password are correct, because I'm using another container with another image and is sending the mail.

Checked also the docs here:

• https://docker-mailserver.github.io/docker-mailserver/v13.1/examples/use-cases/forward-only-mailserver-with-ldap-authentication/#building-a-forward-only-mail-server
• https://docker-mailserver.github.io/docker-mailserver/v13.1/config/advanced/mail-forwarding/relay-hosts/

services:
  mailserver:
    image: ghcr.io/docker-mailserver/docker-mailserver:latest
    container_name: mailserver
    # Provide the FQDN of your mail server here (Your DNS MX record should point to this value)
    hostname: smtp-relay.gmail.com
    ports:
      - "25:25"
      - "587:587"
      - "465:465"
    volumes:
      - ./docker-data/dms/mail-data/:/var/mail/
      - ./docker-data/dms/mail-state/:/var/mail-state/
      - ./docker-data/dms/mail-logs/:/var/log/mail/
      - ./docker-data/dms/config/:/tmp/docker-mailserver/
      - /etc/localtime:/etc/localtime:ro
    environment:
      - PERMIT_DOCKER=network
      - ENABLE_SPAMASSASSIN=0
      - SMTP_ONLY=1
      - ENABLE_FETCHMAIL=0
      - ENABLE_POP3=
      - ENABLE_CLAMAV=0
      - RELAY_HOST=smtp-relay.gmail.com
      - RELAY_PORT=587
      - RELAY_USER=myuse@gmail.com
      - RELAY_PASSWORD=myAPPpassw
    cap_add:
      - NET_ADMIN # For Fail2Ban to work
    restart: always

[polarathene]

hostname: smtp-relay.gmail.com

You are not Gmail, do not pretend to be.

This is the FQDN for your DMS instance to be reached (via those published ports). It's not related to your email accounts that you'll send mail as, but it is what DMS will greet other mail servers as, such as when connecting to Gmail.

You could try just using mail or any other single label hostname.

• Should be ok if you don't have any actual DNS configured to connect an FQDN to DMS (in which case, your ports may not be necessary to publish, or you could publish to only be reachable via the Docker host instead like `"127.0.0.1:25:25").
• Another option is a reserved TLD that cannot be publicly resolved like .test, so hostname: mail.example.test.

---

RELAY_HOST=smtp-relay.gmail.com

You are relaying from DMS back into itself, due to the your hostname misconfiguration.

[albgen]

Using a hostname: "mail" the container does not even start.

Using a hostname: ".test" same as above...

Using a hostname: "mail.example.test" got the initial error "451 4.3.0 Error: queue file write error"

any other ideas?

[casperklein]

I think your problem is related to the operating system, Windows is not supported. You could try running DMS on a linux VM (I recommend Debian) and check if the problem is gone.

[polarathene]

Using a hostname: "mail" the container does not even start.

Ah ok, that is due to:

docker-mailserver/target/scripts/startup/check-stack.sh

Lines 26 to 36 in 0889b0f

	function _check_hostname() {
	_log 'debug' 'Checking that hostname/domainname is provided or overridden'
	_log 'debug' "Domain has been set to ${DOMAINNAME}"
	_log 'debug' "Hostname has been set to ${HOSTNAME}"
	# HOSTNAME should be an FQDN (eg: hostname.domain)
	if ! grep -q -E '^(\S+[.]\S+)$' <<< "${HOSTNAME}"; then
	_dms_panic__general 'Setting hostname/domainname is required'
	fi
	}

Using a hostname: ".test" same as above...

I meant mail.example.test, that should work fine? (our tests use that)

---

Using a hostname: "mail.example.test" got the initial error "451 4.3.0 Error: queue file write error"

Ok, so the write error message you can try google:

• This suggests it's related to Amavis
• This one notes the issue eventually was related to bare domain config (eg: example.test and using example.test mail accounts, or the equivalent of mail.example.test for both). But then also... amavis.

So I would suggest try disabling Amavis for now with ENABLE_AMAVIS=0. If that works great!

Getting it to work with Amavis enabled would be best to start with clean mail-state volume. You need to ensure that you restart DMS correctly, because Amavis config is additive (bug) which is likely to cause problems each time the config is duplicated from improper container restart.

---

However, for a null mailer (SMTP_ONLY=1):

• I'm not sure if it makes sense for you to have Amavis enabled? Or even port 25 in use for that matter.
• You should be sending mail to DMS to be relayed via the submission(s) ports (465/587), from your services configured with credentials (username + password), which should bypass Amavis anyway 🤔

For additional clarification, setup email add (or whatever you use to provision mail accounts in DMS) are not tied to your sender address (except with SPOOF_PROTECTION=1, which presently is not default). So you could have a single account that all your services wanting to submit mail to DMS for relay use, or they can individually have their own accounts (better).

[albgen]

Tried all the suggestions but still nothing. The most stupid test i made was to put a non valid password and i get the same error 451 4.3.0 Error: queue file write error so it seems it does not even try to connect to gmail...

[albgen]

I think your problem is related to the operating system, Windows is not supported. You could try running DMS on a linux VM (I recommend Debian) and check if the problem is gone.

tried on a Rocky linux running docker and it worked...

[albgen]

btw, regarding running it on Windows: if i use volumes instead bind mount on the docker-compose files, i don't get the queue error and the message is queued. On the logs however i see the following:

The difference between windows and linux is that on linux it connects to the tls port instead on windows, is trying to connect to port 25