Reddit, forums, unable to find a clear resolution, as users post to change .htaccess and others discuss changing the swag config or proxy configs, etc
The output of your Nextcloud log in Admin > Logging:
[no app in context] Error: in_array(): Argument #2 ($haystack) must be of type array, null given
GET /ocs/v2.php/apps/files_sharing/api/v1/sharees?format=json&itemType=file&search=&perPage=20&lookup=false
at 2022-10-23T13:14:50+00:00
/config/www/nextcloud/lib/private/AppFramework/App.php - line 156:
OC\AppFramework\Http\Dispatcher->dispatch()
/config/www/nextcloud/lib/private/Route/Router.php - line 301:
OC\AppFramework\App::main()
/config/www/nextcloud/ocs/v1.php - line 62:
OC\Route\Router->match()
/config/www/nextcloud/ocs/v2.php - line 23:
require_once("/config/www ... p")
Caused by TypeError: in_array(): Argument #2 ($haystack) must be of type array, null given
/config/www/nextcloud/apps/circles/lib/Collaboration/v2/CollaboratorSearchPlugin.php - line 101:
in_array()
/config/www/nextcloud/lib/private/Collaboration/Collaborators/Search.php - line 72:
OCA\Circles\Collaboration\v2\CollaboratorSearchPlugin->search()
/config/www/nextcloud/apps/files_sharing/lib/Controller/ShareesAPIController.php - line 228:
OC\Collaboration\Collaborators\Search->search()
/config/www/nextcloud/lib/private/AppFramework/Http/Dispatcher.php - line 217:
OCA\Files_Sharing\Controller\ShareesAPIController->search()
/config/www/nextcloud/lib/private/AppFramework/Http/Dispatcher.php - line 126:
OC\AppFramework\Http\Dispatcher->executeController()
/config/www/nextcloud/lib/private/AppFramework/App.php - line 156:
OC\AppFramework\Http\Dispatcher->dispatch()
/config/www/nextcloud/lib/private/Route/Router.php - line 301:
OC\AppFramework\App::main()
/config/www/nextcloud/ocs/v1.php - line 62:
OC\Route\Router->match()
/config/www/nextcloud/ocs/v2.php - line 23:
require_once("/config/www ... p")
Added the following lines as per documentation, did not resolve the concern
Your web server is not properly set up to resolve “/.well-known/webfinger”.
Your web server is not properly set up to resolve “/.well-known/nodeinfo”.
Your web server is not properly set up to resolve “/.well-known/caldav”.
Your web server is not properly set up to resolve “/.well-known/carddav”.
I have resolved the errors so far and now attempting to finally use the updater to update to the latest build
Each instance I tried failed
first time it failed, I could not return and access the webGUI would get update in process
So I restored my backup and was able to get back into the webGUI again. Checked over permissions on the file shares and attempted to re-run the updater.
I pull docker image containers quite regularly, and the NC version stays on the same release regardless of the docker container I noticed. I thought the same thing, updating the container would grab the most recent version of NC which it was not.
I confirmed the same result with another person with the same setup I often speak with that has NC docker deployed, and that user was also out of date.
They were on a more recent built of NC so they didn’t encounter the concerns I had as of recent, it’s quite the troubleshooting event to update to the latest version of NC I must say, takes some time and digging.
I got past the updater concern tonight, it was a ‘strange fix’ to say the least.
The next concern is regarding the SQL database, but I may have a handle on that, its regarding how NC sets the db schema, which is a known bug in versions of SQL
it uses a COMPRESSED row format which causes NC updates not to process and requires to be changed to DYNAMIC
May break on it for tonight and work some more on it tomorrow.
My apologies I missed responding to that.
NC is pulling from the repository: linuxserver/nextcloud
I hope that’s the information you were looking for?
If you need me to dig down some more, please feel free to ask and I’ll have another peak
I noticed that when running the updater, it looks like it pulls incremental updates as I couldn’t get to the same one as the other user I was speaking with… on the first pass that is. However I am halted at a db concern to address first before the next CU installs for NC.
Ah… okay, that’s not the official Docker image. It says on hub.docker.com that it’s version 25, but who knows.
Their doc actually does say to use the web GUI updater. You may need to get in touch with them for help. I don’t know how to troubleshoot their modified Docker image.
Yes, it was a fluke I found this out as I attempted to log into NC with another device and got a warning prompt, and said the exact same thing (as I update docker images regularly). Then the rabbit hole of concerns began
I appreciate all the feedback, slowly but surely I hope I can resolve the SQL concern with the NC db and move forward getting this up-to-date.
Ah gotcha. lscr.io was on that web link you posted earlier, perhaps I misunderstood.
Oh for sure, I am going to be steady as she goes and get up to snuff with this docker image and version of NC and relax once I am good and this is done and out of the way. I ran into many concerns with attempting to fix this update. The nice thing is the backup and restore process of the docker takes seconds if the instance breaks within processing any update changes which is SUPER handy.
Hope I make it past these hiccups and now knowing how the docker container is separate from the NC updates, I will be sure to setup a maintenance schedule for this along with the other tasks.
The updater fix was ridiculous, basically at the step I was stuck at you have to keep hitting ‘retry’ for little bits of data to process. It seemed to be a bug or permission thing that was causing it
/var/www/html/updater/updater.phar/lib/Updater.php on line 871
which equates to in my case
data/updater-xxxxxxxxxx/downloads/nextcloud/lib/l10n
if you keep hitting retry (took me approx. 40x times) and you have that data directory open while doing it, you see files and folders processing (disappearing from the DIR) and eventually the update process’s. so weird!
I thought I was at the finish line, until the db halted me processing the installation of the update, I seem to be in a better spot then I was and at least have some take-aways from this whole update process!
going to post this as ‘resolved’ for the time being as this concern itself was not resolved, and moved to another db instance that may work better with NC moving forward