I’m running NC 33.0.3 (latest), and all worked perfectly for weeks. Then, yesterday, I upgraded the Linux system, rebooted, and now I can’t access the web gui any more. My browsers, no matter which one I try, tell my they can’t access the server, while at the same time all synchronising apps on all my devices still are doing their job: I can synch files, I can synch my podcasts, I can even access my calendars and my contacts.
And I can even access my files using FolderSync on my phones.
It is just the browsers, that can not connect to the server.All of my browsers, on any of my devices!
What could be the source of this problem? I did not find any related or helpful error message in any log file. Any help will be appreciated. Here are some information about the server:
Armbian Armbian 26.2.1 trixie
nginx 1.26.3
mariadb 11.8.6
php 8.4
SSL-cert by Letsencrypt, valid till end of June.
Just tell my which additional information you need in particular trying helping my to solve this mystery. Many thanks in advance!
Like a non-existant domain, or a server that is shutdown? You can ping the server from a client computer?
When you try to access your server, do you get any new logs on the server (you should get something in the access.log or error.log)? If there is nothing at all, something might block the connection? Strange that it is on all devices, if not it might be a browser plugin (security plugin).
It is a local setup in your local network? And you always access through a public domain (letsencrypt certificate)? On the browser and the apps? Just thinking if you are on a system on a local network, and you changed your address, the redirect on your router might not be up to date.
So just some ideas, to figure out if it is network or Nextcloud related.
The information requested in the support template (that you seem to have removed from your post).
Precisely what occurs when you access the web UI and your installation method for starters. Also possibly things like web console, etc. See the template.
Thanks for your reply! My time is a bit limited at the moment, but I will provide more information in the next days. Meanwhile, I might have a clue to what’s going on:
My /etc/nginx/sites-available/nextcloud includes the use to a port different from 443:
server {
listen 12345 ssl; # Just an example!!!
server_name myserver.name
…
}
Apps and all of my browsers then used the address https://myserver.name:12345, and this has been working flawlessly until a few days ago, no complaining by any app or browser.
My guess is the latest update to nginx has made a change in how the server reacts to different ports. Now, if I tell my browser to go to https://myserver.name:12345 the address bar changes to just https://myserver.name, without the provided port number, and the browser can’t connect.
As a quick and dirty solution I simply added port 443, and suddenly my browsers can connect again:
server {
listen 12345 ssl; # Just an example!!!
listen 443 ssl;
server_name myserver.name;
…
}
It would be interesting to find out why just the browsers (any browser!) are effected, while other apps still are working without adding port 443 to the servers listening.