'Interner Serverfehler' since Update

Hi,

Today at 12:35 my Nextbox started to upgrade Nextcloud.

Since then I get the error, when I try to open the nextbox website. All connecting apps give me an http 500 error, too.

What I tried already:

Reboot of the nextbox

I have ssh access. But i did nothing special there. I noticed the admin page was complaining about some column format, but think this has nothing to do with the problem.

Is this update error already known, is nitrokey working on this?

I may send some logs, if you say which ones you need.

Regards, Thamos

i think i made a little progress: Even occ got me an exception. the message said something about custom_apps/spreed/… . So i moved the spreed folder to spreed.broken. Now php occ apps:list is working again. It lists me the spreed.broken app as disabled (as i would expect.) Somehow spreed got installed by itself again, it is now listed as active again with working occ.

Now the Nextcloud was in maintenance mode. After an hour time, looking into the logs, the problem solved itself :slight_smile: now.

(PS, i used the ssh access and worked inside the docker image.)

Hi Thamos,

do you mind sharing a bit more detailed what you did with occ?

i am also handing in that error, i managed to get on ssh but even php seems not to be installed … oO

Hi,

I entered the docker environment:

sudo docker exec -it -u www-data nextbox-compose_app_1 bash

Called occ to list its apps

php occ app:list

This threw an exception. I don’t have the exact error trace anymore, but it complained about 'custom_apps/spreed/’

You may adapt if some other app seems to be the cause.

What i would try now is

php occ app:disable spreed

What i did was moving the folder custom_apps/spreed to custom_apps/spreed.broken

Then i exited the container, and did sudo shutdown -r now.

After that i waited some hours. The spreed package got new installed by itself. And everything started working again. So i cleaned up by removing the spreed.broken directory

Does this help you?

Good luck,

Thamos

1 Like

ah, thanks. so i entered docker, but its a different trace.

An unhandled exception has been thrown:
Error: Failed opening required ‘/var/www/html/custom_apps/richdocuments/vendor/composer/autoload_real.php’ (include_path=‘/var/www/html/3rdparty/pear/archive_tar:/var/www/html/3rdparty/pear/console_getopt:/var/www/html/3rdparty/pear/pear-core-minimal/src:/var/www/html/3rdparty/pear/pear_exception:/var/www/html/apps:/var/www/html/custom_apps’) in /var/www/html/custom_apps/richdocuments/vendor/autoload.php:20
Stack trace:
#0 /var/www/html/custom_apps/richdocuments/composer/autoload.php(10): require_once()
#1 /var/www/html/lib/private/legacy/OC_App.php(115): require_once(‘/var/www/html/c…’)
#2 /var/www/html/lib/private/AppFramework/Bootstrap/Coordinator.php(80): OC_App::registerAutoloading(‘richdocuments’, ‘/var/www/html/c…’)
#3 /var/www/html/lib/private/AppFramework/Bootstrap/Coordinator.php(49): OC\AppFramework\Bootstrap\Coordinator->registerApps(Array)
#4 /var/www/html/lib/base.php(678): OC\AppFramework\Bootstrap\Coordinator->runInitialRegistration()
#5 /var/www/html/lib/base.php(1194): OC::init()
#6 /var/www/html/console.php(28): require_once(‘/var/www/html/l…’)
#7 /var/www/html/occ(33): require_once(‘/var/www/html/c…’)
#8 {main}

i figure it is richdocuments in my case. Disabling is not working, again a trace pops up. i now renamed the folder and will reboot - keep you updated in a few hours - thanks for the comprehensive guide!

yes, worked - after a short time in mantainance mode it now is up again.

Still, not why i hoped for when i bought it.

well, i am back again on purple LEDs … this time php occ app:list is not throwing an error. @Thamos any chances you can throw me some ideas what do look at?
Box in general reachable but stays on purple (or falls back shortly after going to green) after pressing the reset pin for one time a short period of time

/solved: probably the nextcloud had a reboot cycle or something and occ stumbled over the richdocuments.old folder i created / renamed richdocuments before.

What helped identifying? tail -n 200 /var/log/nextbox.log with some big red errors :smiley:

Hey,

Great — thanks for solving the problem and sharing your solution. I had a good instinct when I cleaned up the folders; that saved me some trouble. I get the feeling the occ tool is very brittle when it comes to custom apps. That seems like a poor design choice in Nextcloud, but I don’t know enough about Nextcloud to complain — I’m just happy when it works. :smiley:

Greetings

Vielleicht kann einer von Nitrokey ( @daringer oder @jan.knittel ) erklären wie es immer wieder zu solchen Problemen kommt und wie das gelöst werden könnte. So wie ich es verstanden habe sollten eigentlich die custom apps deaktiviert werden bevor das Update durchgeführt wird damit es nach dem Update eben nicht zu solchen Problemen kommt. Vielleicht wäre ein Watchdog für den Update-Prozess sinnvoll?