I’ve set up a vserver at a hoster with debian 12 and installed ngninx (apt install ngninx).
Then I’ve configured a dyndns service to access my server, the server has IPv4 and IPv6 adresses:
server1.dyn1.test (expample, not working )
The domain has only the IPv6 adress.
My homeserver is only accessable via IPv6. I’ve set up another dyndns server (other provider) to access the server:
server2.dny2.test
I’ve created the symlink from sites-enabled to the sites-available config file.
in the nginx.conf I’ve added
include /etc/nginx/sites-enabled/*;
Wenn I try to open http://server1.dyn1.test I always get the nginx welcome page.
Tried to figure out what’s wrong via chatGPT, but everything it suggested was OK.
any idea what I should check? Could it be the problem, that I access nginx via IPv4 and it should forward the traffic to an IPv6 only server?
Hey @Merlin123! First things first, can you try installing NGINX from our repository instead of the Debian repository? The Debian 12/bookworm build is four years old at this point and the version disparity makes troubleshooting harder.
In so far as your issue, I just want to double check you have tried to reload NGINX’s config using nginx -s reload? In addition, you mention your servers have IPv6 addresses. However, your NGINX config is set up for IPv4, not IPv6. Our docs cover how to enable it.
If neither of these suggestions prove to be useful, can you share your entire NGINX config using nginx -T or replicate your environment in https://tech-playground.com/playgrounds/nginx/? This will make it much easier to help you out.
I’ve edited your answer to remove some potential sensitive data. Hope you don’t mind! I also made an edit to my previous reply so you might have missed it whilst replying – the issue might also be due to NGINX being configured to only listen on IPv4 whilst your servers are configured to only use IPv6.
Can you also share your access and error logs? They should be under /var/log/nginx.
Ok. Reinstalled nginx, same behaviour.
IPv6: What if the computer trying to open the webpage can’t use IPv6? I’m setting up this server to be able to access my IPv6 pages with an IPv4 only client. Isn’t this possible?
Playground: Never used it before. I could paste my nginx.conf, but where should I paste my server configs?
You would need to have your first server be able to listen on IPv4, and then proxy to the one listening on IPv6. Per your first comment my understanding is that the server1.dyn1.test domain is only set to use the IPv6 address from your server?
If you visit http://ipv6-server.ipv64.de directly I assume you can see your site just fine? Your NGINX config should be proxying to the server no matter what if you hit ipv4-server.duckdns.org directly?
yes, from my laptop it’ working.
and a
curl -6 http://ipv6-server.ipv64.de
on my ipv4 server
returns the right content, so it should be accessable without any issues
Based on everything you have been sharing and your logs, the config should really work just fine. Everything seems to be pointing to your NGINX config not being reloaded. You should be seeing something along the lines of the following in your error log
2025/08/05 21:05:10 [notice] 19#19: signal process started
2025/08/05 21:05:10 [notice] 8#8: signal 1 (SIGHUP) received from 19, reconfiguring
2025/08/05 21:05:10 [notice] 8#8: reconfiguring
Instead of running nginx -s reload, you could also try reloading the service altogether. If that doesn’t work either, I would really need a reproducible environment to offer further help. You can probably use some public endpoints for your proxy target and you should be able to use localhost as your NGINX/IPv4 server.
points to my internal network. There is another nginx running to route the traffic from serveral wizards-tower.ipv64.de subdomains to my internal server and docker container.
So I had the idea that the welcome page is delivered from my internal nginx…
In the browser I enter http://server1.duckdns.org/
And this URL is forwarded to my internal nginx… And it didn’t know this domain…
I’ve changed the config of my internal nginx so that it listens to both domains
But after a restart the behaviour was the same
Than I saw that there was a default config in sites-enabled. After deleting this symlink it worked.
What I need to test:
I have several services which should be accessable via IPv4, so I will create more domains. Then I will change my internal ngix and add the additional doamins. Hope this works, too.
Thanks a lot for your help!
And sorry that I’ve forgotten to mention the second nginx. It just didn’t come to my mind that this could be an issue
The second NGINX instance would indeed explain the behaviour! You should be able to add additional domains to your NGINX config without any further issues by following a similar pattern