Upgrading From Scale Cobia to Dragonfish cause WebUI not working

TrueNAS General
1

Hello.
Yesterday I updated from SCALE Cobia to Dragonfish. I have lost WebUI access, the page does not load. It doesn’t work by http or https. I can access by SSH, SAMBA… but not by WeUI
I need help
Thanks

2

if you have ssh access you could try to restart the truenas middleware with:

sudo systemctl restart middlewared

Also did you clear your browser cache? Sometimes after updates your browser tries do load cached data.

3

Hello
I’ve already tried. But it still doesn’t work. I’m thinking of using the menu option 5. Reset configuration to default

4

Before resetting the config i’d try to restart the machine again. Maybe a process got borked during the update/restart process.

5

Thank you guys.
I was finally able to solve it. Use option 5 of the menu. Import the pool. And all that was left was to create new users and permissions.

6

For future reference of anyone coming to this thread, configuration reset should not be treated as a solution to webui failing to work. In some situations this may have unintended consequences including permanent loss of configuration information, production outage, etc.

If an upgrade to a released version of TrueNAS fails unexpectedly, please gather as much information as possible file a bug ticket, then roll back to the previous working release.

5 Likes
7

Note: I removed “configuration reset” as the approved Solution to this issue.

1 Like
8

Correct. What I have done is like killing flies with cannon shots. I needed to restore the service and it is the only solution I came up with. Obviously it carries a risk and I do not recommend it.

9

##Web UI does not work
Apps, shares and VM’s properly running.
Upgrade from 23.10.2 from WebUI. No issues, regolar reboot.

Tried to restart WebUI via CLI but unsuccessfully.
The /var/log/middleware.log reported a problem on NGiNX not starting.

Tried with systemctl status nginx.service which reports error in configuration files.

systemctl status nginx.service

Apr 25 12:33:33 freenas systemd[1]: Starting nginx.service - A high performance web server and a reverse proxy server…
Apr 25 12:33:33 freenas nginx[3454]: 2024/04/25 12:33:33 [emerg] 3454#3454: invalid port in upstream “:::5902/” in /etc/nginx/nginx.conf:92
Apr 25 12:33:33 freenas nginx[3454]: nginx: configuration file /etc/nginx/nginx.conf test failed
Apr 25 12:33:33 freenas systemd[1]: nginx.service: Control process exited, code=exited, status=1/FAILURE
Apr 25 12:33:33 freenas systemd[1]: nginx.service: Failed with result ‘exit-code’.
Apr 25 12:33:33 freenas systemd[1]: Failed to start nginx.service - A high performance web server and a reverse proxy server.
Apr 25 12:33:37 freenas systemd[1]: nginx.service: Unit cannot be reloaded because it is inactive.

middleware.log

[2024/04/25 12:33:37] (WARNING) middlewared.plugins.service_.services.base.systemd_unit():171 - nginx reload failed with code 1: ‘nginx.service is not active, cannot reload.\n’

10

There is a known open issue which was reported on RC.1 which will be fixed in 24.04.1 release details available here.

VM with IPv6 bind address breaks nginx webserver

If you are in a position to try a nightly build (note these are not official releases) the fix is available in the dragonfish nightly builds.

2 Likes
11

Thanks @ABain. This is exactly my case. I’ll check whether I can disable IPv6 to temporary fix the issue. Otherwise I’ll wait for the 24.04.1

1 Like
12

Hola a todos. Es un poco tarde y hace un par de horas que debería estar ya acostado. Así que perdonad que no traduzca esta entrada. Empiezo a ver algo borroso ya :slight_smile:

Os cuento…

Pues me ha vuelto a pasar. Estaba configurando una maquina virtual. No me di cuenta de que el dispositivo display estaba en IPv6. Y al reiniciar el servidor principal, me ocurrió exactamente lo mismo.

Bien voy a describir los pasos que he dado para recuperar el WebUI sin tener que restaurar la configuración de fabrica:

Accedemos por SSH y comprobamos el estado de nginx.

systemctl status nginx.service

En mi caso me avisa de un error en el puerto 5904. Este puerto lo estaba usando el display de la maquina virtual en cuestión.

Editamos el archivo de configuración de nginx y localizamos el bloque de código en el que aparece el puerto que nos este dando error. En mi caso al ser del display de una máquina virtual decidí comentar el bloque entero.

nano /etc/nginx/nginx.conf

Ahora reiniciamos nginx

systemctl start nginx.service

Si todo ha ido bien podremos acceder al WebUI sin problemas.

Ahora toca localizar la máquina virtual que está usando IPv6 y cambiarlo a IPv4.

Espero que esto sirva de ayuda. Puede que no sea un método muy ortodoxo pero soy nuevo en sistemas y en Linux.