warning: /usr/lib/postfix/sbin/smtpd: bad command startup -- throttling

uniquegch

I am having issues with my mailcow-dockerized. watchdog shows that postfix and dovecot are constantely “going down” by -2 or -10. Also I cannot access my mailboxes in my mailclient, because I am getting Connection refused.
in the logs I can see the warning (shown in the subject line).

any suggestions are welcome.
Thanks

uniquegch

just ran a restore on my mailserver but still system is unusable.

uniquegch

ran a sudo ./update.sh and it had a few images updated. seems that helped, because for now I can access my e-mails and watchdog is not showing any of the health issues in the previous post (for now).

uniquegch

and I am starting having the issues with Dovecot, postfix and also CertificateCheck.
This leads to refusing the connection from my mailclient to the mailcow server.

uniquegch

how can fix the issue with dovecot and postfix, which seems to cause a connection refused from my mail client to the server mailboxes

uniquegch

one solution for the above described situation is running the update.sh script. I just executed it and for pretty much each container was an updated image available. I will keep an eye on that and see if the problem with the health status of dovecot and postfix is related with status of the current image.

uniquegch

and the problems are reoccuring and the ./update.sh “trick” does not work either.
and I must assume I am the only one who seems to have that problem I will have to look for another mailserver solution.
Thank you community for your outstanding help.
You are all a blast.

esackbauer

uniquegch Thank you community for your outstanding help.

We can only help if you are sharing your installation basis, which OS and Version, docker and docker compose versions, which hoster or self hosted, IPv6 enabled or not, and not really providing log files for nginx and postfix and watchdog.
So yes, it was a blast guessing what you are running mailcow on 😉

uniquegch

hello esackbauer,

hmm ok and where do I find all those information?
Thanks
regards

esackbauer

uniquegch hmm ok and where do I find all those information?

Sorry, if you really don’t know then you should consider paying for a mailcow installation (e.g. servercow.de)… That is Linux basics and reading the docs:
https://docs.mailcow.email/troubleshooting/debug/

uniquegch

this is a little bit of a longer answer.

well I did some looking around and came up with the following

VPS: 5.10.0-26-amd64 #1 SMP Debian (bullseye) 5.10.197-1 (2023-09-29) x86_64 GNU/Linux
Docker version 24.0.7
Docker Compose version v2.18.1

that VPS I am using only to run mailcow-dockerize

I changed the setting of “… enable_ipv6: true …” to false (in the docker-compose.yml file around line 615).

At this point watchdog shows me that postfix and dovecot are healthy, actually everything is healthy, except when I look with docker containter ps (kinda of see more details below)

Current issue is also the container for unbound
“… mailcow/unbound:1.18 ”/docker-entrypoint.…“ (unhealthy) …”

after running docker container logs (ID unbound) I checked the firewall making sure the port 53 is not blocked / denied, rebootet the server (for $h!t and giggles) and when it was back online postfix and dovecot crapped out on me. and it makes sense because they cannot resolve DNS names (duh unbound not healthy).

Now did some “hammer” tooling and ran.
docker-compose kill
docker-compose up -d

NOTE: all containers have been created but not all have started like postfix, watchdog, acme, netfilter, clamd. and unbound caused an error.

ok so killing seems at bit harsh. so lets be gentle and do
docker-compose stop
docker-compose up -d
ah wonderful it shows all good and unbound heathly
… hmm wait in docker container ps it shows container unbound unhealthy (WTF). But let me check, well watchdog shows everything healthy and all my mailboxes can be opened. So I guess after some much time x the $h!t will hit the fan again.