Preparing indefinitely

meray

OS: Rocky Linux 9.5
Branch: Master 1994b9895bc92cfc83caf0f25123efbb763f8635
PHP-FPM: 1.93
Docker: 28.2.2
SELinux: active

I tried this both with and without IPv6 support _{I don’t have IPv6 in my network, changes are commented}

    # enable_ipv6: true
    ipam:
      driver: default
      config:
        - subnet: ${IPV4_NETWORK:-172.22.1}.0/24
        #- subnet: ${IPV6_NETWORK:-fd4d:6169:6c63:6f77::/64}

also editting resolv.conf to remove the faulty search domain, but going to the server it stays on
preparing for

/
/admin

and 404 for

/SoGo
/debug

the output of docker compose logs --tail=200 php-fpm-mailcow nginx-mailcow provides:

nginx-mailcow-1| 2025/06/02 13:42:18 [error] 15#15: *4 connect() failed (111: Connection refused) while connecting to upstream, client: [client IPv4], server: mail.mydomain.tld, request: "GET / HTTP/1.1", upstream: "fastcgi://172.22.1.6:9002", host: "[mailcow IPv4]"
nginx-mailcow-1| [client IPv4] - - [02/Jun/2025:13:42:18 +0000] "GET / HTTP/1.1" 502 2015 "-" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36"
nginx-mailcow-1| 2025/06/02 13:42:18 [error] 15#15: *4 connect() failed (111: Connection refused) while connecting to upstream, client: [client IPv4], server: mail.mydomain.tld, request: "GET / HTTP/1.1", upstream: "fastcgi://172.22.1.6:9002", host: "[mailcow IPv4]"
nginx-mailcow-1| 2025/06/02 13:42:18 [error] 15#15: *4 connect() failed (111: Connection refused) while connecting to upstream, client: [client IPv4], server: mail.mydomain.tld, request: "GET / HTTP/1.1", upstream: "fastcgi://172.22.1.6:9002", host: "[mailcow IPv4]"
nginx-mailcow-1| [client IPv4] - - [02/Jun/2025:13:42:18 +0000] "GET / HTTP/1.1" 502 2015 "-" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36"

and

php-fpm-mailcow-1  | Waiting for SQL...

alternatingly.

source mailcow.conf ; docker compose exec mysql-mailcow mysql -u${DBUSER} -p${DBPASS} ${DBNAME}
logs me into maria with no errors.

Thanks in advance! 😃

_{As a sidenote.. my network is 10.10.0.0/24, not 172.22.1.0/24 as the the docker-compose.yml would suggest but I’ve seen no instruction to change it so I assume it is a magic number}

meray

I think the sidenote may hold the clue.
When I # ./generate_config.sh

Found Docker Compose Plugin (native).
Setting the DOCKER_COMPOSE_VERSION Variable to native
Notice: You'll have to update this Compose Version via your Package Manager manually!
Press enter to confirm the detected value '[value]' where applicable or enter a custom value.
Mail server hostname (FQDN) - this is not your mail domain, but your mail servers hostname: [mail.mydomain.tld]
Timezone [America/New_York]: [City]
Which branch of mailcow do you want to use?

Available Branches:
- master branch (stable updates) | default, recommended [1]
- nightly branch (unstable updates, testing) | not-production ready [2]
- legacy branch (supported until February 2026) | deprecated, security updates only [3]
Choose the Branch with it's number [1/2/3] 1
Already on 'master'
Your branch is up to date with 'origin/master'.
Generating snake-oil certificate...
----- [SnakeOil omitted]
Copying snake-oil certificate...
Detecting if your IP is listed on Spamhaus Bad ASN List...
Check completed! Your IP is clean

It seems a tad short especially considering mailcow.conf is 300 lines.
I tried running it in proxmox shell to make sure it is not an ssh issue but it persists.

esackbauer

As SElinux is active, have you done this?
https://docs.mailcow.email/getstarted/install/#check-selinux-specifics

meray

esackbauer Yes. I also restarted the whole vm since then.

I actually changed the
IPV4_NETWORK=10.10.0
line to this in mailcow.conf restarted the entire image ^{because whenever I change either mailcow.conf or d-c.yaml, dc up fails with very weird error messages}, but now the entire server is unavailable through ssh or 80.

esackbauer

Why did you change the IPV4_NETWORK? Who told you so?

meray

I don’t know why I can’t edit the garbled response above, but the host is completely unresponsive on ssh and can’t ping other hosts on his network. Verifying ip a | grep ens18 ipadress is still 10.10.0.50/24

[unknown] Nobody, but I’m pretty much out of ideas and since I couldn’t find the documentation on what the magic number there meant I figured it’d be part of the problem

esackbauer

Please leave the default network. It is the docker internal network for mailcow.
You should ONLY change that if it collides with the IP range of your VM.

Any changes to mailcow.conf will not be active, even after a reboot, until you do
docker compose up -d

I would also suggest you start from scratch, ideally with a minimal Debian Netinst, instead of Rocky Linux. Well it can be “rocky”, as some here have already experienced. Debian and Ubuntu are pretty solid for mailcow.

meray

esackbauer Well I did and now it’s working. What I don’t get is why they hint at rocky linux being a viable alternative… But thank you very much 😃

SOLUTION

Don’t use Rocky Linux. Use DebIan Latest

CLOSED